Assessing mobility in primary brain tumor patients: A descriptive feasibility study using two established mobility tests

Abstract

Background

Patients with primary brain tumors (PBT) face significant mobility issues related to their disease and/or treatment. Here, the authors describe the preliminary utility and feasibility of two established mobility measures, the Timed-Up-and-Go (TUG) and Five-Times Sit-to-Stand (TSS) tests, in quickly and objectively assessing the mobility status of PBT patients at a single institution’s neuro-oncology clinic.

Methods

Adult patients undergoing routine PBT care completed the TUG/TSS tests and MD Anderson Symptom Inventory-Brain Tumor module (MDASI-BT), which assessed symptom burden and interference with daily life, during clinic visits over a 6-month period. Research staff assessed feasibility metrics, including test completion times/rates, and collected demographic, clinical, and treatment data. Mann–Whitney tests, Kruskal–Wallis tests, and Spearman’s rho correlations were used to interrogate relationships between TUG/TSS test completion times and patient characteristics.

Results

The study cohort included 66 PBT patients, 59% male, with a median age of 47 years (range: 20–77). TUG/TSS tests were completed by 62 (94%) patients. Older patients (P < .001) and those who were newly diagnosed (P = .024), on corticosteroids (P = .025), or had poor (≤80) KPS (P < .01) took longer to complete the TUG/TSS tests. Worse activity-related (work, activity, and walking) interference was associated with longer TUG/TSS test completion times (P < .001).

Conclusions

The TUG/TSS tests are feasible for use among PBT patients and may aid in clinical care. Older age, being newly diagnosed, using corticosteroids, poor (≤80) KPS, and high activity-related interference were associated with significant mobility impairment, highlighting the tests’ potential clinical utility. Future investigations are warranted to longitudinally explore feasibility and utility in other practice and disease settings.

Primary brain tumors (PBT) are typically associated with a poor prognosis and significant symptom burden.^1,2 Mobility issues, including impaired gait and balance, are commonly observed in neuro-oncology clinical practice.^3–6 Other motor deficits, such as weakness and impaired proprioception, can further compromise quality of life, functional independence, and safety.⁷ These deficits may arise from the impact on critical neuroanatomy of the tumor and/or related treatment.⁸ Tumor treatments can compound tumor-associated symptoms, leading to worsening or addition of new symptoms.

Regardless of the cause, PBT patients across the disease trajectory are often highly symptomatic and face functional limitations. As many PBT patients report spending a significant portion of their lives feeling ill and unable to perform usual activities, these symptoms and limitations can impact crucial areas of one’s life, such as returning to work.^9–11 Symptom burden also has documented clinical relevance. For instance, patient-reported activity-related interference, including challenges with working or walking, is predictive of tumor progression.¹²

Given the prevalence of mobility-related issues in PBT patients and the clinical significance of capturing mobility information, neuro-oncology providers require tools to assess their patients’ mobility during clinic visits. Currently, several tests enable providers to capture such data, most prominently the 6-Minute Walk Test (6MWT).¹³ However, quicker and more subjective measures, such as the Karnofsky Performance Status (KPS) scale, are commonly used to assess patients’ mobility. KPS has been plagued by poor interrater reliability and a weak value in predicting prognosis when performance status is deemed “good” (KPS ≥ 90%).^13–15 Additionally, while more objective mobility measures such as the 6MWT have been used to assess PBT patients’ physical functioning, their use is limited by restricted space in the clinic.⁶

In a recent study by Dulfikar and colleagues (2021), the authors described their clinical use of the Timed-Up-and-Go (TUG) test to assess physical functional capacity of patients with glioma prior to adjuvant radiation therapy.¹⁶ This mobility test, along with the Five-Times Sit-to-Stand (TSS) test, has previously been identified as clinically relevant in patient populations with other neurological diseases, such as Parkinson’s disease,^17,18 multiple sclerosis,¹⁹ and stroke,²⁰ and has potential for telehealth application. The TUG and TSS tests have been demonstrated to be important in predicting fall risk²¹ and decline in global health and functioning,²² which makes them intriguing potential tests in neuro-oncology. Ultimately, there is a clinical need in neuro-oncology for a quick, accurate, and objective mobility measure. Therefore, the authors conducted an exploratory, cross-sectional study at a single-institution to preliminarily describe the utility and feasibility of leveraging the TUG and TSS tests, two established mobility assessments, in routine neuro-oncology clinical practice.

Materials and Methods

Study Cohort Assembly

Patients were recruited from The University of Texas MD Anderson Cancer Center’s (MDACC) Brain and Spine Center, where they were receiving routine PBT care, to participate in a large, descriptive study exploring aspects of symptom burden in the PBT patient population. Research staff reviewed electronic health records (EHRs) to determine patient eligibility, which included: 1) being ≥18 years of age; 2) able to read, speak, and write English; and 3) capable of providing informed consent. Exclusion criteria included: 1) neurological complications of systemic cancer; and 2) inability to complete self-report surveys due to cognitive deficits, as determined by the clinical team. During a clinical evaluation, research staff approached PBT patients and their caregivers, confirmed study eligibility, and obtained the patient’s written informed consent before collecting data.

For statistical analyses, enrolled patients were divided into three groups: 1) newly diagnosed: within one month of a diagnostic neurosurgical procedure or diagnostic neuroimaging if a surgical intervention could not be performed; 2) recurrence: recurrent disease and initiated on a new treatment regimen; and 3) long-term survivors: at least three years out from an initial diagnostic neurosurgical procedure in which a PBT diagnosis was made. Additionally, patients were grouped according to treatment phase: newly diagnosed, on active treatment, or in follow-up care. Cross-sectional data collection was conducted over a 6-month period. The study’s procedures were reviewed and approved by the MDACC Institutional Review Board (IRB) and conform to established clinical research standards.

Study Measures

Sociodemographic and clinical variables.

—Patients completed baseline forms at study entry. From EHRs, research staff extracted relevant demographic, clinical, and treatment characteristics, including sex, age, race, ethnicity, patient group, treatment phase, World Health Organization (WHO) tumor grade, tumor type, KPS, body mass index (BMI), tumor location, medication use (corticosteroids and anticonvulsants), and tumor recurrence status.

The description of KPS ≤80 as “poor” was made based on previous work from our research team which demonstrated differences in patient rating of both symptom severity and functional limitations (in terms of self-reported activities of daily living) at this cutoff value.¹ Additionally, those who reported corticosteroid use completed the Dexamethasone Symptom Questionnaire-Chronic (DSQ-C), an 18-symptom and 4 open-ended item patient self-reported measure of the incidence and severity of side effects associated with their corticosteroid use. Cumulative steroid dose was calculated as dose in milligrams (mg) multiplied by duration (ie, days from self-reported corticosteroid start date to mobility test date).

Mobility Tests

Timed-Up-and-Go (TUG) test.

—The TUG test is designed to assess mobility and balance. During test administration, the patient sits in a standard armchair and, on the instructor’s command of “Go,” rises and walks 3 meters (~10 feet) at a comfortable and safe pace using their regular footwear and a walking aid, if required (see Figure 1). The patient then turns, walks back, and sits down in the chair. A stopwatch records the total time spent completing the test, beginning at the “Go” instruction and stopped when the patient returns to the initial seated position.²³

Figure 1.

The Timed-Up-And-Go (TUG) test is designed to assess mobility and balance. The patient first sits in a standard armchair. On the test instructor’s command of “Go,” the patient rises from the chair and walks 3 meters (m). The patient then turns, walks back to the chair, and sits down. Timing, using a stopwatch, begins at the instruction of “Go” and stops when the patient returns to their initial seated position.

Importantly, the TUG test has demonstrated select measures of interrater reliability²⁴ and validity among elderly patients^25,26 and those with Parkinson’s disease^27,28 and stroke.^25,29

(TSS) test.

—The TSS test is designed to assess functional lower extremity strength.^30,31 During test administration, the patient sits in an armless chair with their back straight, feet shoulder-width apart, hands on opposite shoulders, and wrists crossed and held against the chest (see Figure 2). On the instructor’s “Go” signal, the patient rises to a full standing position and then returns to the initial seated position. The patient completes this test sequence five times. Research staff monitor the patient’s performance to ensure proper form. Similar to the TUG test, a stopwatch records the total time spent completing the TSS test, beginning at the “Go” instruction and stopped when the patient returns to their initial seated position after the fifth trial.

Figure 2.

The Five-Times Sit-To-Stand (TSS) test is designed to assess functional lower extremity strength. The patient sits in the middle of an armless chair and is instructed to keep their back straight and feet approximately shoulder-width apart. The patient crosses their arms (as depicted in the inset image) and on the instructor’s signal of “Go,” rises to a full standing position and then returns to the initial seated position. The patient is instructed to complete this test sequence 5 times. Research staff monitor the patient’s performance to ensure proper form. Timing, using a stopwatch, begins at the instruction of “Go” and stops when the patient returns to their initial seated position after the fifth trial. Reprinted figure in the public domain from the Centers for Disease Control and Prevention, 30-second chair stand assessment handout, Atlanta, GA. https://www.cdc.gov/steadi/pdf/STEADI-Assessment-30Sec-508.pdf. Accessed on 6 June 2021.

The TSS test has demonstrated select measures of safety, reliability, and validity in older intensive care unit patients at discharge³² and community-dwelling elderly patients.³³ Additionally, TSS test administration has reportedly high interrater reliability.³⁴

Brain Tumor-Related Symptoms and Interference

Patients also completed the MD Anderson Symptom Inventory-Brain Tumor module (MDASI-BT), which consists of 22 items assessing brain tumor-related symptom severity and 6 items assessing interference with daily life.¹ The symptom severity items measure 6 underlying symptom factors, including affective, cognitive, focal neurologic deficit, generalized disease, treatment-related, and gastrointestinal-related symptoms. These symptom items are rated “at their worst” on a 0–10 scale (0 = not present; 10 = bad as can be imagined). The interference items capture activity-related (ie, work, activity, and walking) and mood-related (ie, mood, enjoyment of life, and relations with other people) interference data. Higher scores indicate higher symptom severity or interference. For statistical analyses, symptoms rated ≥5 and interference ≥2 were considered “moderate-to-severe.”

Feasibility Metrics and Utility

Feasibility was determined by the number/percent of patients who could complete one or both mobility tests. An additional feasibility metric was the time required for each test to be completed. Importantly, the feasibility of using the TUG/TSS tests assessed in the present study was a secondary, exploratory objective of a larger study conducted at MDACC analyzing changes in PBT patients’ symptom burden. For this larger study’s primary objectives, a sample size of 50 newly diagnosed PBT patients was deemed necessary to detect an effect size of 0.40 between time points with power of 0.80 using a two-tailed test with an alpha of 0.05. This difference between time points was not reported in the present manuscript, which had no a priori thresholds. Utility was defined as how the two mobility tests distinguish known groups who would be predicted to take longer (eg, elderly patients, those with poor KPS, and those on corticosteroids).

Statistical Analyses

Descriptive statistics, including means, medians, standard deviations, and percentages, were used to characterize the patient cohort and report TUG/TSS test completion times. For analyses, BMI was divided into three categories: normal weight (18 ≤ BMI < 25); overweight (26 ≤ BMI < 29); and obese (BMI ≥ 30). TUG/TSS test completion times were evaluated for normality. Because the distributions of these scores were not normal, differences based on patient demographic, clinical, and treatment characteristics were evaluated nonparametrically using the Mann–Whitney U test for dichotomous variables and Kruskal–Wallis test for variables with ≥3 categories. Relationships between patient demographic, clinical, and treatment characteristics and TUG/TSS test completion times were assessed using Spearman’s rho correlations. Effect sizes were reported as correlation coefficients and eta squared (η ²). The level of statistical significance was set at P < .025, adjusting for multiple comparisons of the two mobility tests.

Results

Study Cohort

Baseline demographic and clinical data for the 66 PBT patients in the study cohort are displayed in Table 1. Patients were predominantly obese (41%), White (86%) men (59%) with a median age of 47 years (range: 20–77). Half of the cohort had a WHO grade IV PBT (50%), with 47% of patients diagnosed with glioblastoma according to criteria from the WHO 2016 classification of central nervous system (CNS) tumors. Regarding phase along the disease trajectory at the time of study enrollment, 29% were newly diagnosed, 27% were in recurrence, and 44% were in the long-term survivorship period. Most patients had a good (≥90) KPS (73%; range: 60–100) and a frontal (39%) or temporal (33%) lobe tumor. Additionally, while the majority of patients were taking anticonvulsants (n = 56, 30 of whom were on levetiracetam), far fewer were taking corticosteroids (n = 23, 22 of whom were on dexamethasone).

Table 1.

Patient Demographic, Clinical, and Treatment Characteristics at Baseline

Patient characteristics (N = 66)	n (%)
Sex
Female	27 (41%)
Male	39 (59%)
Age, median (range)	47 (20–77)
Race
White	57 (86%)
Black/African American	3 (5%)
Asian	2 (3%)
American Indian/Native Alaskan	1 (2%)
Missing data	3 (5%)
Ethnicity
Hispanic	9 (14%)
Karnofsky performance status (KPS)
≤80 (Poor)	18 (27%)
60	5 (8%)
70	4 (6%)
80	9 (14%)
≥90 (Good)	48 (73%)
90	26 (39%)
100	22 (33%)
Body mass index (BMI)
Normal weight (18 ≤ BMI < 25)	25 (38%)
Overweight (26 ≤ BMI < 29)	14 (21%)
Obese (BMI ≥ 30)	27 (41%)
Patient group
Newly diagnoseda	19 (29%)
Recurrenceb	18 (27%)
Long-term survivorsc	29 (44%)
Treatment phase
Newly diagnosed	19 (29%)
On treatment	25 (38%)
In follow-up	22 (33%)
WHO tumor grade
I	1 (2%)
II	14 (22%)
III	18 (27%)
IV	33 (50%)
Tumor type d
Glioblastoma (Grade IV)	31 (47%)
Astrocytoma (Grade I, II, or III)	20 (30%)
Oligodendroglioma (Grade II or III)	7 (11%)
Othere	8 (12%)
Tumor location f
Frontal lobe	26 (40%)
Temporal lobe	22 (34%)
Parietal/occipital lobe	11 (17%)
Multiple lobes involved	6 (9%)
Medication use
Corticosteroid	23 (35%)
Dexamethasone	22 (96%)
Prednisone	1 (4%)
Anticonvulsant	52 (79%)
Levetiracetam	30 (58%)
Otherd	22 (42%)
Tumor recurrence
Yes	29 (45%)
No	36 (55%)

^aPatients within one month of a diagnostic neurosurgical procedure or diagnostic neuroimaging if surgical intervention could not be performed.

^bPatients who had experienced a recurrence and were initiated on a new treatment regimen.

^cPatients who were at least three years from their initial diagnostic neurosurgical procedure in which a primary brain tumor diagnosis was made.

^dTumor type is according to the WHO 2016 classification of central nervous system tumors.

^e“Other” tumor types represented in this patient cohort include: oligoastrocytoma, gliosarcoma, diffuse glioma, high-grade glioma, infiltrating glioma, and pilocytic astrocytoma. “Other” anticonvulsants include: phenytoin (2), carbamazepine (1), sodium valproate (3), phenobarbital (1), lamotrigine (1), and other (18).

^fAdditionally, 1 patient had a tumor in the corpus callosum.

Feasibility Findings

Nearly all (94%) patients completed the two mobility tests. Four patients could not complete the TUG/TSS tests because they were unable to stand (n = 3) or walk without assistance (n = 1). Therefore, 62 PBT patients were considered in our analyses. Patients took 5–40 s to complete the TUG test, with an average time of 13 s, and 4–67 s to complete the TSS test, with an average time of 15 s. Almost all patients (97%) completed the five required TSS test trials, with one patient completing only one trial and another completing just three trials.

Patient Demographics

Table 2 includes select associations between patient characteristics and mobility test completion times. Age impacted the time patients took to complete the mobility tests. TUG/TSS test completion times correlated positively with age (TUG: r = 0.53, P < .001; TSS: r = 0.50, P < .001). No statistically significant differences in mobility test completion times were identified based on any other patient demographic, such as sex or BMI.

Table 2.

Associations Between Select Patient Characteristics and Mobility Test Completion Times

Patient Characteristic		Timed-Up-And-Go						Five-Times Sit-To-Stand
		N	Median	Mean	SD	Sig.	Effect sizea	N	Median	Mean	SD	Sig.	Effect sizea
Sex	Male	37	10.0	12.5	6.5	0.270	0.14	37	12.0	13.7	5.9	0.494	0.09
	Female	25	11.0	13.6	7.2			25	13.0	17.4	14.2
Karnofsky performance status (KPS)	Poor (≤80)	14	17.5*	19.1	8.1	<0.001	0.52	14	17.5*	20.9	12.4	0.005	0.36
	Good (≥90)	48	10.0*	11.2	5.2			48	12.0*	13.5	8.9
Body mass index (BMI)	Normal weight	22	10.0	10.9	3.3	0.167	0.03	23	12.0	13.5	5.8	0.579	−0.02
	Overweight	14	9.0	12.6	6.3			14	11.0	19.6	19.0
	Obese	26	12.5	14.9	8.6			25	14.0	14.2	4.6
Patient group	Newly diagnosed	17	11.0	14.5	7.7	0.051	0.07	18	15.0	17.1	7.0	0.032	0.08
	On treatment	23	12.0	13.9	7.6			22	12.0	16.3	15.1
	In follow-up	22	9.0	10.8	4.5			22	12.0	12.5	4.6
Treatment phase	Newly diagnosed	17	11.0	14.5	7.7	0.198	0.02	18	15.0*	17.1	7.0	0.024	0.09
	Recurrence	17	11.0	13.4	7.9			16	11.0*	14.5	14.3
	Long-term survivors	28	10.0	11.8	5.4			28	12.0	14.3	9.3
WHO tumor grade	I–II	15	10.0	10.9	3.4	0.031	0.09	15	11.0	12.3	4.2	0.155	0.03
	III	17	10.0	11.5	5.8			16	12.0	14.4	11.4
	IV	29	12.0	15.0	8.1			30	14.0	17.0	11.6
Tumor location (lobe)	Frontal	22	10.0	10.6	3.2	0.375	0.02	23	12.0	12.3	4.0	0.401	0.02
	Temporal	22	11.0	14.8	9.7			21	13.0	15.5	10.4
	Parietal/occipital	11	11.0	13.5	5.9			11	13.0	18.1	17.1
Corticosteroid use	Yes	6	14.0*	13.7	3.8	0.025	0.28	6	15.5	17.3	8.2	0.058	0.24
	No	41	10.0*	12.6	7.7			41	12.0	13.8	8.3
Anticonvulsant use	Yes	50	11.0	13.6	7.2	0.067	0.23	49	13.0	15.9	11.1	0.306	0.13
	No	12	9.0	10.4	3.9			13	12.0	12.3	5.2
Tumor recurrence	Yes	28	12.0	13.8	7.1	0.23	0.15	27	13.0	16.0	13.7	0.526	0.08
	No	33	11.0	12.4	6.6			34	12.5	14.6	6.4

Note: all median, mean, and standard deviation (SD) values reported in this table are mobility test completion times measured in seconds. Additionally, 1 patient had a tumor in the corpus callosum and was excluded from the univariate test for tumor location.

*Bolded text indicates statistically significant differences in the median mobility test completion times (ie, P < .025).

^aEffect sizes: correlation coefficient (r; sex, Karnofsky Performance Status (KPS), corticosteroid use, anticonvulsant use, and tumor recurrence); eta squared (η ²; Body Mass Index (BMI), patient group, treatment phase, WHO tumor grade, and tumor location (lobe)).

Clinical and Treatment Characteristics

Patient performance status impacted mobility test completion times, with those with poor (≤80) KPS taking longer than those with good (≥90) KPS to complete the TUG (median = 17.5 s vs. 10.0 s, r = 0.52, P < .01) and TSS tests (median = 17.5 s vs. 12.0 s, r = 0.36, P < .01). Patients’ phase along the disease trajectory impacted only TSS test completion times, with newly diagnosed patients taking longer than those with recurrent disease (median = 15.0 s vs. 11.0 s, η² = 0.09, P = .024). In contrast, there was no statistically significant difference in test completion times by treatment phase, history of tumor recurrence, WHO tumor grade, or tumor location. While PBT patients with frontal lobe tumors had the shortest completion times on both tests, this difference did not reach the designated level of statistical significance. Importantly, 1 patient had a tumor in the corpus callosum and was excluded from the univariate test for tumor location.

Medication Usage: Corticosteroids and Anticonvulsants

Twenty-three patients (35%) reported using corticosteroids, including dexamethasone (n = 22) and prednisone (n = 1). The mean dexamethasone dose taken was 7 mg (median = 4 mg, SD = 5 mg, range: 2–18 mg), while the one patient on prednisone took a 20 mg dose. A statistically significant difference in TUG test completion times was found based on corticosteroid use. Those on corticosteroids took longer to complete the TUG test than those not currently on corticosteroids (median = 12.0 seconds vs. 10.0 s, r = 0.28, P = .025).

Among patients who reported corticosteroid use and completed the DSQ-C (n = 21), there was no statistically significant association between DSQ-C score and TUG (r = −0.23, P = .342) or TSS (r = 0.16, P = .492) test completion times. However, a larger cumulative steroid dose was associated with longer TUG test completion time. For the 12 patients (19%) who provided a start date for their corticosteroid use, cumulative steroid dose positively correlated with TUG (n = 11, r = 0.80, P = .003) but not TSS (n = 10, r = 0.62, P = 0.059) test completion times. Finally, no statistically significant differences in test completion times were found based on anticonvulsant drug use.

Symptom Burden and Interference

The top-5 most commonly reported moderate-to-severe symptoms were fatigue (24%), difficulty remembering (24%), feeling drowsy (24%), disturbed sleep (20%), and pain (17%). Additionally, many patients reported moderate-to-severe overall (38%), activity-related (38%), and mood-related (35%) interference. Those reporting higher overall symptom interference took significantly longer to complete the TUG (r = 0.50, P < .001) and TSS (r = 0.53, P < .001) tests. This relationship was particularly strong for activity-related interference (TUG: r = 0.58; TSS: r = 0.56).

Comparison to Normative Data

Previous studies have described cutoff time values to identify patients at risk for falling. The cutoff values for the TUG test are >13.5 s for community-dwelling adults and >11.5 s for patients with Parkinson’s disease. Based on these values for TUG completion times, 24–34% (depending on the comparison population) of our patient sample would be at risk for falling. In addition, cutoff time values for the TSS test are >12 s for community-dwelling adults and >16 s for patients with Parkinson’s disease. Based on these cutoff time values for the TSS test, 18–50% (depending on the comparison population) of our patient sample would be at risk for falling.

When comparing the mobility test completion times from this study’s patient sample to data from previously published studies, it was found that PBT patients took, on average, longer to complete both the TUG and TSS tests than community-dwelling adults.²⁵ Additionally, PBT patients took less time, on average, on the TUG and TSS tests compared to patients from a study of patients with Parkinson’s disease.¹⁷ Conversely, when comparing our sample with a different Parkinson’s disease patient group,¹⁸ PBT patients took more time, on average, to complete both mobility tests. Finally, patients also took longer to complete the TUG test than those in another glioma patient population recently described by Dulfikar and colleagues (2021) in which patients took a median of 7 s to complete the test.³⁵

Discussion

This descriptive, cross-sectional study outlines the preliminary feasibility of using two established mobility assessments in a neuro-oncology clinic. To assess feasibility, we used metrics of test completion times and rates. Research staff administered the TUG/TSS tests with a 94% completion rate. In considering salient patient characteristics, we identified that older age, being newly diagnosed, using corticosteroids, having poor (≤80) KPS, and reporting high activity-related interference were associated with mobility impairment, highlighting the tests’ potential clinical utility. Therefore, our preliminary data suggest a possible role for using the TUG/TSS tests in routine neuro-oncology clinical practice.

The Impact of Age, Sex, and Body Mass Index (BMI) on Mobility Test Completion Times

Both TUG and TSS test completion times increased directly with age. Ibrahim and colleagues (2017) report similar findings related to age in their cohort of older adults.³⁶ Gomes and colleagues (2015) also report increasing TUG times with older age.³⁷ Among healthy elderly individuals, Mangano and colleagues (2020) used data from a wearable inertial sensor to support the postulation that longer TUG test completion times among older adults result from difficulty with the turning portion of the test.³⁸ Elderly individuals are also more likely to be frail,³⁹ contributing to impaired mobility and highlighting the importance of incorporating comprehensive geriatric assessments into routine neuro-oncology practice.⁴⁰

Interestingly, we found no statistically significant differences in mobility test completion times based on sex or BMI. In contrast, in their sample of community-dwelling adults, Bohannon and colleagues (2007) demonstrated that weight and BMI were significantly correlated with TSS test completion times.⁴¹ In their regression models, they showed that age and BMI explained 43.7% of the variance in TSS test completion times, prompting their conclusion that TSS test times should be considered in light of patients’ age and BMI. Given that increased BMI can disrupt balance and postural control, we expected to detect findings similar to those reported by Bohannon and colleagues in our cohort, which had a fairly even distribution of BMI (see Table 1). However, we did not find such significant associations, suggesting the need for further investigations into the relationship between BMI and mobility test completion times in the PBT patient population.

Karnofsky Performance Status and Mobility Test Completion Times

Patients with poor (≤80) KPS took longer to complete both mobility tests than those with good (≥90) KPS. This finding was to be expected as KPS is an indication of a patient’s performance status and is associated with functional mobility. Therefore, our study suggests congruency between a subjective mobility measure and objective mobility assessments. Such congruency has been described previously in a retrospective study of cancer patients seeking geriatric evaluation where poor (≤80) KPS and TUG test performance were directly associated with a history of falls.⁴²

Assessing a patient’s performance status is clinically important and provides prognostic information. It has been established that poor (≤80) KPS is associated with an increased risk of mortality among patients with CNS tumors.⁴³ Similarly, in non-CNS cancers, functional mobility assessments like the TUG test appear to be associated with prognosis, with longer test completion times associated with poor prognosis.^44–46 Thus, our study results suggest that, similar to what has been reported in other cancer populations, subjective performance status, and objective mobility tests, when used in tandem, provide comprehensive clinical information on a patient’s general health status.

The Impact of Phase Along the Disease Trajectory and Tumor Location and Grade on Mobility Test Completion Times

Patients’ phase along the disease trajectory was associated with differential performance on the TSS test, with newly diagnosed patients taking longer than those with recurrent disease. This finding contrasts with previous studies demonstrating that patients with recurrent disease report more motor deficits and worse physical functioning than newly diagnosed patients.^47,48 However, in these two studies, Osoba and colleagues (1997 and 2000) did not utilize objective mobility tests. The longer TSS test completion times found among newly diagnosed patients in our cohort may be accounted for by perioperative deficits in these patients that improve after their tumor burden is diminished and subsequent treatment is initiated.

Despite this difference between newly diagnosed and recurrent patients, a history of recurrence did not impact TUG/TSS test completion times. This finding could be explained by those long-term survivors with no history of recurrence who were grouped with newly diagnosed patients. Long-term survivorship is heterogenous, with patients potentially falling into distinct symptom cohorts with differing physical functioning and symptom complaints.⁴⁹

In neuro-oncology, functional deficits result both from the tumor’s anatomical location and associated treatments.⁸ However, we did not observe a statistically significant difference in test completion times based on tumor location. While patients with frontal lobe tumors had slightly shorter test completion times, the difference was not statistically significant. These findings may be explained by compensation of function; while frontal lobe areas (eg, primary motor and premotor cortices) are important for motor functioning, other brain regions associated with gait and mobility may compensate for frontal lobe tumor-related motor impairments.

The Relationship Between Corticosteroid and Anticonvulsant Drug Use and Mobility Test Completion Times

Corticosteroid use was reported in 35% (n = 23/66) of our cohort, with most patients on dexamethasone (n = 22, 96%). Patients on corticosteroids took longer to complete the TUG test than those not on such medications. In neuro-oncology, corticosteroids are frequently prescribed as adjuncts to tumor-directed therapies and to manage peritumoral edema. Importantly, corticosteroids have been independently associated with significantly shorter overall survival and frequently result in severe side effects.^50,51 As outlined previously, patients in our study who were on corticosteroids completed the DSQ-C instrument, self-reporting side effects related to their corticosteroid use. This specific measure was selected as members of our group previously demonstrated its utility and associated signs and symptoms in a PBT patient population.⁵²

Notable side effects of corticosteroids include steroid-induced myopathy, musculoskeletal complications, and proximal muscle weakness.⁵³ Importantly, PBT patients typically remain on corticosteroids for long periods of time, often until death.⁵⁴ Studies have reported an increased likelihood of side effects with prolonged or higher doses of corticosteroid use.⁵¹

Therefore, with functionally significant and highly prevalent side effects related to corticosteroid use, particularly with long exposure periods and/or high dosages, we expected to see increased TUG/TSS test completion times among those PBT patients with high DSQ-C scores. Interestingly, we found no statistically significant relationship between these variables. However, when considering the relationship between cumulative steroid dose and mobility test completion times, larger cumulative steroid doses were associated with significantly longer TUG test completion times. As first outlined by Arvold and colleagues (2017), while previous studies of corticosteroid use in neuro-oncology have reported idiosyncratic side effects, proximal muscle weakness increases in a predictable “corticosteroid dose x time” dependent manner. Our findings support Arvold’s claim, highlight the clinical utility of the TUG test in detecting proximal muscle weakness among PBT patients, and underscore the need for steroid-sparing therapies in neuro-oncology.

Additionally, 52 patients (n = 52/66, 79%) were on anticonvulsants, with no significant impact on mobility test completion times. These null findings were unsurprising given that, for PBT patients prone to seizures, anticonvulsants mitigate the uncontrolled electrical activity that could compromise mobility and balance.⁵⁵ Future clinical studies assessing mobility among PBT patients should further characterize the anticonvulsant drugs and dosages used, considering, for example, the impact of taking multiple or high doses of anticonvulsant medications on mobility test completion times and rates.

Symptom Burden and Interference

Higher symptom interference, particularly activity-related interference, was associated with longer mobility test completion times. This finding was to be expected as the patient-reported outcome of activity-related interference encompasses activity and walking. Furthermore, previous studies have identified that patient-reported symptoms are associated with functional status.⁵⁶ Additionally, the symptoms of fatigue, difficulty remembering, feeling drowsy, disturbed sleep, and pain commonly reported by patients in the present study are similar in severity to those symptoms observed in other PBT patient samples.²

The Potential for Assessing Mobility During Telehealth Visits: A New Frontier

The coronavirus disease-2019 (COVID-19) global pandemic significantly altered clinical practice and impaired access to health care. However, telehealth visits frequently replaced in-person consultations, offering an essential opportunity to provide remote medical care to diverse, underserved patient communities during this public health emergency.

Importantly, we believe telehealth has promise in cancer far beyond the COVID-19 pandemic. Future studies should consider how feasible using the TUG/TSS tests are in telehealth. Certainly, caretakers will be required to guide patients as they complete the mobility tests and provide safety and aid should a fall occur. Unlike the TUG test or 6MWT, which require extensive physical space, the TSS test may be particularly well-suited for telehealth given the limited space required for test administration, which also ensures patients remain in the computer’s frame of view for the test administrator during the assessment.

Limitations

This study has several limitations worth noting. First, our predominantly White cohort was derived from only one large, tertiary academic medical center in the United States (ie, MDACC). Thus, findings have potentially limited generalizability. As only 27% of patients (n = 18/66) had a poor (≤80) KPS, larger studies considering the performance of these individuals on the mobility tests are warranted. Given that MDACC has hundreds of oncology clinical trials with typically stringent study inclusion criteria, it is unsurprising that the majority of our cohort had a good (≥90) KPS.^57,58

Additionally, patients were all either newly diagnosed, in recurrence, or long-term survivors. Therefore, certain patient populations, for instance those within one year of diagnosis without disease recurrence, were not considered. This convenience sampling issue further limits the generalizability of our findings. Certainly, some clinical factors also occur more commonly in specific patient groups. For instance, in this study, newly diagnosed patients were more likely to be on corticosteroids. However, a robust analysis of differences between treatment groups was limited by our small sample size. Longitudinal research will be crucial to consider possible interactions with demographic and other salient patient characteristics. Such longitudinal work will also highlight how use of the TUG/TSS tests is impacted as patient conditions change throughout the disease trajectory. In the present study, our limited sample size resulted in comparisons of small groups, although we attempted to address this concern with use of multiple comparison adjustments, tests for normality, and appropriate inferential tests. Effect sizes should be considered when interpreting the results presented in this manuscript.

Our study cohort also had a median age of 47 years. Since the TUG/TSS tests were originally designed for frail elderly individuals, these tests may not have been sensitive enough for use among nongeriatric patients. Again, longitudinal research is needed to understand the clinical utility and sensitivity of these mobility tests across the lifespan. Also, we were unable to explore associations with the individual items of the MDASI-BT in this small study. Future studies exploring associations with specific symptoms will be undertaken with a larger sample size, using multiple comparison adjustments, and with a priori hypotheses as to which specific symptoms to evaluate.

Lastly, it is important to consider that PBT patients are a unique population, essentially suffering from both cancer and a neurological disease.⁵⁹ Thus, in our study, it was difficult to parse out the relative impact on mobility issues of the cancer, neurological symptoms, and associated treatments.

Conclusions

Our preliminary study sought to determine if using the TUG/TSS tests was feasible among PBT patients in a neuro-oncology clinic and whether larger studies of these tests are warranted. Secondary objective results from our larger study indicate that the TUG/TSS tests are feasible and clinically sensitive to important patient characteristics, such as age, KPS, phase along the disease trajectory, corticosteroid use, and activity-related interference. These mobility assessments require little equipment, training, and financial expense, and have been shown to be predictive of fall risk in other populations. These tests may assist, along with a comprehensive neurological exam and subjective measures, in guiding clinicians as they select treatments for and evaluate the rehabilitation and homecare needs of their PBT patients.

Future investigations are warranted to explore feasibility in multi-center studies, over longer periods of time, and in other cancer patient populations. Additional work will also be needed to elucidate the role of these mobility assessments in comprehensive neurological exams. Ultimately, our findings underscore the need for longitudinal data collection to gauge how mobility changes throughout the disease trajectory for individual PBT patients and inform development of interventions that can improve mobility status and quality of life.

Funding

This work was supported [in part] by the Intramural Research Program of the National Institutes of Health (NIH), National Cancer Institute, Center for Cancer Research, Neuro-Oncology Branch.

Conflict of interest statement

All authors declare no conflicts of interest.