Daily functioning in glioma survivors: associations with cognitive function, psychological factors and quality of life.

Aim: Understanding and supporting quality of life (QoL) and daily functioning in glioma patients is a clinical imperative. In this study, we examined the relationship between cognition, psychological factors, measures of health-related QoL and functioning in glioma survivors. Materials & methods: We examined neuropsychological, self-reported cognition, mood and QoL correlates of work and non-work-related daily functioning in 23 glioma survivors, and carried out linear models of the best predictors. Results & conclusion: A total of 13/23 participants were working at the time of enrollment. The best model for worse work-related functioning (R2 = .83) included worse self-reported cognitive function, depression, loneliness and brain tumor symptoms. The best model for worse non-work-related functioning (R2 = .61) included worse self-reported cognitive functioning, anxiety, sleep disturbance and physical functioning. Neuropsychological variables were not among the most highly correlated with function. Worse cognitive, particularly self-reported and psychosocial outcomes may compromise optimal functioning in glioma survivors.

Malignant gliomas are neuroepithelial tumors with recent US incidence rates of approximately 4.8 per 100,000 [1]. Lower grade glioma (LGG) survivors in particular benefit from more favorable outcomes compared with other tumor types and comprise approximately 40% of all glioma diagnoses [2]. LGG tends to occur in younger ages compared with other brain tumors and have better overall survival rates with some reports of survival up to 20 years [2,3]. Given the relatively higher likelihood of survival for years after treatment, understanding how to support quality of life (QoL) and daily functioning in glioma survivors, especially LGG, is a clinical imperative receiving growing attention [4-7].

Resuming normal activities in glioma survivorship is critical to QoL [8]. Unfortunately, the survivorship phase is often etched with disruptive cognitive and functional impairments; 30–50% of survivors encounter significant cognitive and functional impairment after treatment and it is one of the most concerning outcomes for patients [9-12]. Even though it is less likely that patients will suffer from more severe cognitive impairment seen in faster growing tumors, even mild cognitive changes can have a profound impact on higher-level functional goals [13,14]. For instance, approximately half of LGG survivors return to work after treatment [15]. Cognitive and other changes threaten glioma survivors ability to carry out complex daily life activities and limit the ability to return to work and fill meaningful and important roles [4-7,16,17].

To help elucidate key factors related to functioning in glioma survivorship, including LGG, we conducted a study of cognition, psychological factors, QoL and functioning in glioma survivors. We grounded our approach in the more developed extant literature available in the parallel field of non-CNS cancer survivors and assembled an extensive neuropsychological and psychosocial battery of measures. In this preliminary report of the first sample of recruited patients, we aimed to understand contributing factors to work and non-work-related daily functioning.

Materials & methods

Sample & study design

This is a cross-sectional observational study of cognitive and psychosocial functioning in glioma survivors. Participants with a history of glioma were recruited from the UCLA Neuro-Oncology clinic. Eligible participants were over the age of 18, had at least 5th grade proficiency in English, and had completed treatments (i.e., surgery, radiation and/or chemotherapy) at least 6 months prior to enrollment. They also must have had a Karnofsky performance status >60%, be able to undergo MRI and pregnant women were excluded.

Enrolled participants underwent same-day research evaluations added on to clinical follow-up appointments. Participants completed a self-report questionnaire framed in biopsychosocial theory, as well as a targeted battery of neuropsychological tests, detailed below. This study also involved extended neuroimaging with fMRI in addition to clinical structural scans, which are not included in the current report. This study was approved by the UCLA Institutional Review Board and all participants signed informed consent.

Neuropsychological & psychosocial measures

Dependent variables

Daily functioning was measured using the work productivity and activity impairment (WPAI) instrument [18], which yields continuous scores for how much the participant’s brain tumor affected: work productivity; and non-work ability or functioning outside the workplace. Scores range from 1 to 10 and are converted into percentages, with higher scores indicating more functional impairment.

Independent variables

The neuropsychological test battery was selected in consideration of expert recommendations for assessment patients with a primary brain tumor, International Cognition and Cancer Task Force recommendations and authors’ clinical experience [19,20]. The battery included the following measures: test of premorbid functioning [21]; The Hopkins Verbal Learning Test – Revised [22]; The Brief Visuospatial Memory Test – Revised [23]; The Trail-Making [24,25]; Wechsler Adult Intelligence Scale-IV [26] Coding and Digit Span subtests; The Connor’s Continuous Performance Test-II [27]; The Golden Stroop [28]; Verbal fluency – FAS and animal naming [29]; The Boston Naming Test [30]; and The Rey–Osterrieth complex figure [31]. Raw scores were transformed into standard scores using published normative data.

Subjective cognitive functioning was assessed using the functional assessment of cancer therapy-cognitive function (FACT-Cog) version 3 [32,33]. The FACT-Cog yields four sub scores derived from items using a 5-point Likert scale to rate impairment: perceived cognitive impairment; perceived cognitive ability; comments from others; and impact on QoL. Depression symptoms were assessed using the Beck depression inventory-2nd edition [34], anxiety using the Beck Anxiety Inventory [35] and trauma symptoms were evaluated using the revised Impact of Event Scale [36]. We measured sleep disturbance using the Pittsburgh Sleep Quality Index [37], with missing values replaced by the mean of completed items [38]. Fatigue was assessed using the multidimensional fatigue symptom inventory – short form [39].

Health-related QoL was measured using the FACT-brain, a global QoL measure for cancer survivorship that also incorporates items specific to the brain tumor population [40]. This scale yields the subscales physical well-being, social/family well-being, emotional well-being, functional well-being, the brain cancer subscale and a total. In order to avoid confounds with the outcome (i.e., functioning), we did not include the functional well-being or total in analyses.

Analytic approach

Simple demographic information was summarized for the sample (see Table 1). We also compared those who were still working to those who stopped working on neuropsychological and self-report outcomes using t-tests (see Table 2). For descriptive purposes, we identified those with cognitive impairment defined as having more than two tests >-2 z-score compared with normative data, consistent with International Cognition and Cancer Task Force guidelines accounting for the number of tests in the battery [19,41]. We took a two-step approach toward modeling WPAI work productivity (among those still working) and WPAI non-work ability (among the entire sample. First, preliminary correlations were conducted to identify relevant correlates to include in the models (see Table 3). We then conducted two exploratory linear regression models of WPAI Work Productivity and WPAI non-work ability to evaluate the magnitude of effects (see Table 4). Since this is a small sample we minimized the number of predictors to four in each model [42]. We also examined model diagnostics to ensure <5% of standardized residuals were greater than 1.96. All analyses were conducted using IBM SPSS Statistics for Mac software, v. 24. (IBM Corp., NY, USA).

Table 1.

Demographic characteristics of study population.

Characteristic		Total sample (n = 23)
		n	Percent
Age, years
Mean	44.26
SD	12.24
Education
Mean	16.04
SD	1.77
Time since surgery, years
Mean	7.60
SD	(5.06)
Intelligence quotient (IQ)
Mean	104.96
SD	12.33
Gender
Male		15	34.8
Female		8	65.2
Race
White, non-Hispanic		21	91.3
African–American		1	4.3
Asian		1	4.3
Employment status
Full or part-time		10	43.5
Not employed		13	56.5
Tumor location
Left frontal		5	21.7
Left occipital		1	4.3
Left parietal		1	4.3
Left temporal		4	17.4
Right frontal		6	26.1
Right frontal temporal		1	4.3
Right occipital		1	4.3
Right parietal		2	8.7
Right temporal		1	4.3
Subcortical		1	4.3
IDH mutant
Yes		15	65.2
No		3	13
N/A		4	17.4
Missing		1	4.3
Radiation therapy
Yes		21	91.3
No		2	8.7
Chemotherapy
Yes		21	91.3
No		2	8.7
Chemotherapy type
PCV, tamoxifen		1	4.3
TMZ/PCV/CCNU/procarb/vincristine		1	4.3
TMZ, adjuvant		18	78.3
Velcade/TMZ, adjuvant		1	4.3
N/A		2	8.7
Cancer diagnosis, revised
AA		5	21.7
AOA		3	13
AOG		2	8.7
Anaplastic mixed glioma		1	4.3
GBM		4	17.4
LA		4	17.4
LO		2	8.7
PPA		1	4.3
PXA		1	4.3

AA: Anaplastic astrocytoma; AOA: Anaplastic oligoastrocytoma; AOG: Anaplastic oligodendroglial gliomas; CCNU: Lomustine; GBM: Glioblastoma; LA: Low grade astrocytoma; LO: Low grade oligodendroglioma; PCV: Procarbazine + lomustine + vincristine chemotherapy; PPA: Pilocytic astrocytoma; PXA: Pleomorphic xanthoastrocytoma; SD: Standard deviation; TMZ: Temozolomide (Temodar).

Table 2.

Psychosocial factors between diffuse glioma survivors currently working or not working.

Variable	Working (n = 13)	Not working (n = 10)	Total sample (n = 23)
Employment change post brain tumor, n (%)
No	11 (85)	3 (30)	14 (61)
Yes (reduced, personal time, stopped)	2 (15)	7 (70)	9 (39)
Age, mean (SD)	42.85 (14.69)	46.10 (8.49)	44.26 (12.24)
Gender, n (%)
Male	10 (77)	5 (50)	15 (65)
Female	3 (23)	5 (50)	8 (35)
Years of education, mean (SD)	16.23 (1.301)	15.80 (2.30)	16.04 (1.77)
IQ, mean (SD)	105.08 (14.69)	104.8 (9.16)	104.96 (12.33)
Time since surgery (years), mean (SD)	8.08 (5.78)	6.96 (4.14)	7.60 (5.06)
Impaired on two or more neuropsychological tests ≤-2 z-score, n (%)	7 (54)	5 (50)	12 (52)
BDI-II, mean (SD)	8.08 (8.751)	11.6 (9.08)	9.61 (8.87)
BAI, mean (SD)	5.15 (5.94)	12.3 (9.61)	8.26 (8.38)†
FACT-Br, mean (SD)
Physical well-being	24.92 (4.07)	23.6 (4.74)	24.35 (4.32)
Social/family well-being	21. 69 (6.41)	20.7 (5.06)	21.26 (5.75)
Emotional well-being	21 (2.92)	17.9 (3.9)	19.65 (3.65)†
Functional well-being	20.15 (8.05)	18.7 (5.06)	19.52 (6.81)
Brain cancer subscale (n = 22)	74.96 (14.75)	59.78 (17.98)	68.75 (17.49)†
Total (n = 22)	162.73 (26.35)	139.11 (29.85)	153.07 (29.62)
FACT-Cog, mean (SD)
Perceived cognitive impairments	54.38 (18.19)	37.5 (18.05)	47.04 (19.67)†
Comments from others	13.92 (2.93)	13 (3.53)	13.52 (3.16)
Perceived cognitive abilities	18.08 (7.66)	12.1 (7.23)	15.48 (7.91)
Impact of quality of life (n = 22)	14.42 (2.43)	10.3 (5.31)	12.55 (4.43)†
MFSI-SF, mean (SD)
General scale	6.27 (7.12)	5.1 (4.47)	5.76 (6.01)
Physical scale	2.54 (3.97)	5.1 (5.08)	3.65 (4.57)
Emotional scale	3.69 (3.90)	5.75 (5.63)	4.59 (4.73)
Mental scale	5.58 (5.17)	10.1 (6.83)	7.54 (6.24)
Vigor scale	15.08 (5.42)	12.8 (6.22)	14.09 (5.76)
Total score	3 (22.06)	13.25 (23.43)	7.46 (22.74)
PSQI, mean (SD)
Global (n = 18)	4.86 (4.11)	6.71 (3.45)	5.58 (3.87)
Global with avg missing items (n = 22)	5.73 (5.59)	6.58 (3.27)	6.08 (4.71)
Loneliness scale total (n = 22), mean (SD)	4.23 (1.42)	5.11 (1.96)	4.59 (1.68)
Self-efficacy scale total, mean (SD)	50.92 (10.02)	39.1 (17.07)	45.78 (14.49)†
Impact of event scale total, mean (SD)	10.92 (10.5)	20.8 (15.12)	15.22 (13.37)

p < 0.05.

BAI: Beck Anxiety Inventory; BDI-II: Beck Depression Inventory – second edition; FACT-Br: The functional assessment of cancer therapy – brain; FACT-Cog: The functional assessment of cancer therapy – cognitive function; IQ: Intelligence quotient; MFSI-SF: Multidimensional fatigue symptom inventory – short form; PSQI: The Pittsburg Sleep Quality Index; SD: Standard deviation.

Table 3.

Correlations with work productivity and activity impairment productivity and work productivity and activity impairment ability.

Demographic and clinical variables	WPAI productivity (n = 13)	WPAI ability (n = 23)
Age	0.236	-0.074
Gender	-0.357	-0.147
Education, years	-0.311	0.329
IQ	-0.269	0.036
Time since surgery§	0.364	-0.002
Chemotherapy (yes/no)	0.345	0.193
Neuropsychological tests, normed
Letter fluency total	-0.447	0.124
Animal naming total	-0.421	-0.132
BNT-2 total	-0.366	-0.163
BVMT-R total recall	-0.394	0.029
BVMT-R delayed recall	-0.368	0.049
Coding total	-0.261	-0.426†
Digit span total	-0.325	-0.273
HVLT-R total recall	-0.624†	-0.276
HVLT-R delayed recall	-0.467	-0.151
HVLT-R recognition§	-0.367	-0.198
Rey-O copy§	-0.401	-0.409
Stroop word reading	-0.360	-0.128
Stroop color naming	-0.263	-0.223
Stroop color-word	-0.391	-0.308
Stroop interference§	-0.451	-0.114
TOPF	-0.269	0.036
Trails A time	0.131	-0.068
Trails B time	-0.144	-0.346
Impaired on two or more neuropsychological tests ≤-2 z-score, n (%)	0.479	0.374
Self-report measures
BDI-II raw	0.827‡	0.576‡
BAI Raw	0.468	0.609‡
FACT-Br physical well-being§	-0.621†	-0.583‡
FACT-Br social/family well-being§	-0.469	-0.446†
FACT-Br emotional well-being	-0.269	-0.469†
FACT-Br functional well-being§	-0.250	-0.533‡
FACT-Br brain cancer subscale	-0.612†	-0.506†
FACT-Cog perceived cognitive impairments	-0.760‡	-0.542‡
FCAT-Cog perceived cognitive abilities	-0.851‡	-0.564‡
PSQI global§	0.587†	0.721‡
Loneliness scale total	0.818‡	0.414
Impact of event scale total§	0.607†	0.527‡

p < 0.05.

p < 0.01.

Variable significantly skewed (Z >1.96), Spearman’s rank correlation coefficient.

BAI: Beck Anxiety Inventory; BDI-II: Beck Depression Inventory – second edition; BNT-2: Boston naming test, 2nd edition; BVMT-R: Brief visuospatial memory test-revised; FACT-Br: The functional assessment of cancer therapy – brain; FACT-Cog: The functional assessment of cancer therapy – cognitive function; HVLT-R: Hopkins verbal learning test-revised; IQ: Intelligence quotient; PSQI: The Pittsburg Sleep Quality Index; Rey-O: Rey–Osterrieth complex figure test; Stroop: Stroop golden; TOPF: Test of premorbid functioning; WPAI: Work productivity and activity impairment questionnaire.

Table 4.

Linear regression models of work productivity and activity impairment productivity and ability.

DV	Variable	R2	R2adjusted	β	Standard error	Standardized beta	Partial correlation	Part correlation	Variance inflation factor
WPAI productivity		0.832	0.748
	FACT-Cog perceived cognitive abilities			-0.302†	0.120	-0.974	-0.665	-0.365	7.128
	FACT-br physical well-being			-0.209	0.172	-0.359	-0.395	-0.176	4.152
	BDI-II raw			0.075	0.076	0.277	0.329	0.143	3.756
	Loneliness scale total			-1.072	0.717	-0.643	-0.467	-0.216	8.816
WPAI ability		0.610	0.518
	FACT-Cog perceived cognitive abilities			-0.158†	0.073	-0.448	-0.465	-0.328	1.861
	FACT-br physical well-being			-0.238	0.139	-0.386	-0.383	-0.259	2.223
	BAI raw			0.078	0.066	0.246	0.277	0.180	1.861
	PSQI global			-0.085	0.147	-0.146	-0.139	-0.087	2.792

p < 0.05.

BAI: Beck Anxiety Inventory; BDI-II: Beck Depression Inventory – second edition; FACT-Br: The functional assessment of cancer therapy – brain; FACT-Cog: The functional assessment of cancer therapy – cognitive function, v.3; PSQI: The Pittsburg Sleep Quality Index; WPAI: Work productivity and activity impairment questionnaire.

Results

We recruited 24 gliomas survivors for the study. Participants had various glioma diagnoses, predominantly lower grade, including anaplastic astrocytoma ([AA]; 5/23), anaplastic oligoastrocytoma ([AOA]; 3/23), anaplastic oligodendroglial liomas ([AOG]; 2/23), anaplastic mixed glioma (1/23), glioblastoma ([GBM]; 4/23), low grade astrocytoma ([LA]; 4/23), low grade oligodendroglioma ([LO]; 2/23), pilocytic astrocytoma ([PPA]; 1/23) and pleomorphic xanthoastrocytoma ([PXA]; 1/23). Broadly, participants were more likely to be positive for the IDH mutation, with nearly half of participants with frontal tumor location (11/23). Participants were also likely to have received radiation therapy (21/23) and chemotherapy (21/23), with TMZ adjuvant as the most common type (18/23). One participant was excluded for low English fluency detected during testing. The analytic sample included 23 participants, demographic and clinical characteristics displayed in Table 1. In general, the sample was mostly male, White, with an average estimated verbal IQ. At the time of this study, 13 participants were currently working; 11 reported no change in employment and two had reduced their time at work. Ten participants were not working, of which seven reported they had stopped working due to the brain tumor. Table 2 displays comparisons on neurocognitive and psychosocial factors between those currently working and those not working (neuropsychological scores are summarized in Supplementary Table 1). No differences were observed on neurocognitive outcomes between those working and those not working. Compared with working survivors, those not currently working reported higher levels of anxiety, worse perceived cognitive impairment, worse emotional well-being, worse symptoms related to brain cancer and lower sense of self-efficacy (p < .05). Mean WPAI work productivity scores were 18.46% (±23.75) among the 13 working and mean WPAI non-work ability scores were 22.17% (±27.13%), with higher percentages indicating greater impairment.

Preliminary bivariate correlations identified several associations with both WPAI work productivity and WPAI non-work ability scores, see Table 3. Among demographic and clinical variables, none were associated with WPAI work productivity or WPAI non-work ability. Few neuropsychological test scores were significantly correlated with either measure except better learning was associated with higher WPAI work productivity and better attention was associated with higher WPAI non-work ability (p’s <.05). In contrast, several self-report correlates emerged for both measures including mood, physical well-being, brain tumor symptoms, perceived cognitive impairment, loneliness and sleep disturbance. Given the small sample and relatively large effects, we selected the four strongest relationships for each measure for exploratory regression models; of note, we selected the highest of the two mood measures highly correlated with WPAI non-work ability (i.e., the BAI). We also examined part and partial correlations in the models for further scrutiny of unique effects since the strongest associations were in self-report measures.

The model statistics for both WPAI Work Productivity and WPAI non-work ability are presented in Table 4. For WPAI Work Productivity, 83% of the variance was explained by the BDI-II, Loneliness, the FACT-Cog Perceived Cognitive Abilities, and the FACT-Brain Physical Well-Being, F(4,8) = 9.92, p < .01. The FACT-Cog Perceived Cognitive Abilities score was the strongest and only significant predictor in the model, but given the small sample size we also examined the partial and part correlations of the other predictors, which each seem to carry some unique predictive value to the model with small to moderate effect sizes after accounting for the other predictors. The model of WPAI non-work ability was also significant, F(4,17) = 6.65, p < .01, explaining 61% of the variance with measures of subjective cognitive functioning (FACT-Cog Perceived Cognitive Abilities), anxiety (the BAI), sleep disturbance (PSQI) and physical functioning (FACT-brain physical well-being). Again, the FACT-Cog Perceived Cognitive Abilities score was the strongest and only significant factor. Examining partial and part correlations revealed some small to moderate unique effects of the BAI and FACT-brain physical well-being, but the PSQI dropped to negligible effects in the presence of the other factors.

Discussion

Functioning in brain tumor survivorship is an important factor to promote QoL in these patients, a critical clinical mandate [43]. The present study is a preliminary report of functional outcomes among glioma survivors participating in a comprehensive QoL study. Our goal in this report was to begin understanding factors that contribute to work and non-work related daily functioning in glioma survivors, including LGG, which comprised the majority of our sample. In our sample, roughly half of the survivors were not currently employed, consistent with others’ findings [15]. Rates of impairment in work productivity and non-work ability were comparable to those reported in other samples of patients with chronic conditions (i.e., multiple sclerosis) [44]. The majority of those no longer working reported that it was due to the impact a brain tumor has had on their life. At the bivariate level, we found a range of psychological factors associated with both work and non-work related functioning, including objective and subjective cognitive measures, mood, self-reported physical functioning and self-reported loneliness. Linear models of work and non-work related functioning both highlighted perceived cognitive functioning, physical well-being and mood as unique predictors.

Surprisingly, neuropsychological outcomes were not strongly associated with functional outcomes. We found no differences on individual neurocognitive test scores between those currently working and those not working, nor different rates of cognitive impairment. Further, no neurocognitive measure was as strongly associated with self-reported cognition for either work or non-work-related daily functioning. There are high rates of cognitive impairment in glioma patients across their trajectory of treatment and into survivorship, and even in lower grade tumors [14,45,46]. Cognitive impairment is a significant risk to QoL and functioning in patients with glioma [43]. Noll and colleagues' reported that neuropsychological functioning significantly predicted independent functioning among newly diagnosed patients with temporal lobe tumors of mixed type [47]. They found specifically memory and executive function measures predicted functional independence, although self-reported cognition measures were not included. By comparison, our study employed several of the same neuropsychological measures, but our sample was smaller and comprised mostly of glioma survivors well after completing treatment, with varying tumor locations. Our results do not frankly contradict their conclusion about the ecological validity of neuropsychological measures: we did observe some relationships during preliminary correlational analyses consistent with their findings, and the correlation coefficients in many instances are relatively high and of a similar magnitude to the Noll et al. study, if non significant – notably the cognitive impairment variable and memory and executive function measures. Since, this was a small preliminary study, further research will be critical to more fully understand neuropsychological predictors of functioning.

Even in this small sample, though, our preliminary findings do point to a few important lines of inquiry that will require further pursuit with larger samples. Our results highlight a unique role of perceived cognitive function. The significance of perceived cognitive impairment on function in the non-CNS cancer survivorship literature is also gaining attention. Cognitive changes in glioma patients, including LGG, may be more mild than those experienced by patients with other types of brain tumor, as seen in our sample [48]. Even if mild, can significantly compromise higher-level functional goals such as returning to work or resuming family roles [16,17]. While objective neuropsychological testing is critical to characterizing deficits, self-reported cognitive functioning may provide a unique window into the patient's experience and priorities that can help signal need for attention and intervention, such as cognitive rehabilitation.

The rich literature in non-CNS cancer populations similarly portrays functioning be complex, shaped by multiple factors including, but not limited to, cognitive impairment [10,19,49] and emphasizes the importance of self-reported cognitive function [50]. Our findings from this preliminary study are well-aligned with this position, suggesting a cluster of influences on daily functioning including cognitive function and psychological factors. Overall, the results of our study adds a new dimension of appreciating self-reported cognitive challenges in brain tumor survivors as well as the likely contributions of psychological factors and highlights the need for more study.

The need to address psychological needs in brain tumor survivors is intensifying [51,52]. We know that cognitive and functional changes often manifest in the setting of psychosocial risk factors such as the well-described increases in anxiety and depression symptoms in brain tumor and other cancer patients [53-55]. Similar to our findings, Feuerstein and colleagues examined contributing factors to work productivity in brain tumor survivors and also found depressive symptoms to independently contribute [56]. Others have similarly found mood to be a predictor of health-related QoL in recently diagnosed glioma patients [57]. Furthermore, patients’ functional goals and sense of meaning can be altered by brain tumor diagnosis and an anticipated shortened survival [51]. Loughan and colleagues recently found high rates of death-related distress in primary brain tumor patients irrespective of tumor grade [58]. Given the complexity of needs, it is critical to incorporate a biopsychosocial framework in developing models of optimal functioning for these patients, as also advocated by others [5,8,51].

This study has limitations. First, we acknowledge the restrictions of small sample size in terms of statistical power and generalizability. Glioma survivors, especially LGG, are a fraction of all brain tumor survivors, which are a small fraction of all cancer survivors in general. There is also substantial heterogeneity in our study sample including cognitive impairment, time since treatment and tumor location, which could be influencing outcomes and limiting our ability to detect relationships with factors such as cognitive function. Importantly, we also primarily focused on cognitive, psychological and QoL outcomes – it remains important to understand the full nature of potential contributing factors to functioning in this population, such as specific physical impairments and type of work. Despite the limitations of size of the population from which we sampled, UCLA is a tertiary center and draws patients from a wide region, permitting accrual of relatively larger samples than seen in general oncology clinics. Thus, while small by statistical standards, the collected sample is a unique aggregate of data from this population. Additionally, our sample was predominantly White with higher levels of educational attainment and more study in more diverse samples is imperative.

Conclusion

This study lends preliminary evidence to support a multi-factorial approach to providing supportive care in glioma survivorship. Despite favorable outcomes in terms of longevity for LGG survivorship, the cognitive and psychological toll of this experience compromises optimal functioning and it is critical to further study and develop interventions for optimizing functioning and QoL for these patients.

Survival among glioma survivors, especially lower grade, is improving.

However, 30–50% of survivors experience significant functional impairment.

To better support glioma survivors, it is important to understand factors contributing to functional impairment.

This preliminary study examines cognitive, psychological and quality of life predictors of work and non-work-related functional impairment in 23 glioma survivors. A total of 13/23 survivors in our sample were currently working.

Linear models of the strongest correlates of work productivity included self-reported cognitive function, depression, loneliness and brain tumor symptoms, explaining 83% of the variance.

Linear models of the strongest correlates of non-work-related ability included self-reported cognitive function, anxiety, sleep disturbance and brain tumor symptoms.

The results of our preliminary study highlight a multifactorial approach to supporting functioning in glioma survivorship including cognitive, both objective and self-report and psychological factors. Further research in this area with larger samples will be important for developing targeted and effective supportive interventions in this population.

Click here for additional data file.