Abstract
Background
Glioma research has increasingly emphasized quality of life alongside traditional survival metrics, emphasizing functional outcomes, symptom burden, and social reintegration, including the ability to work. While previous studies focused on return-to-work rates, we assessed general work ability as a broader measure of work capacity. We aimed to develop predictive models for general work ability recovery, identify key risk factors, and explore long-term trajectories.
Methods
We conducted a retrospective cohort study of 342 patients with glioma (aged 18–64, WHO Grades 2–4) between March 2010 and December 2018. Work ability and symptoms were assessed using the M.D. Anderson Symptom Inventory-Brain Tumor Module (MDASI-BT), which was administered at months 1 and 3 postoperatively, then every three months up to 12 months, and at longer intervals thereafter. Logistic regression predicted 6-month general work ability recovery, and Cox models identified long-term risk factors. Long-term monitoring was conducted to evaluate the stability of work ability recovery across different WHO grades.
Results
65.2% (223/342) regaining general work ability within 6 months post-surgery. Brain tumor-specific symptoms were stronger predictors of recovery than general symptoms. Predictive models achieved AUCs of 0.78 (pre-surgery) and 0.82 (post-surgery). Long-term monitoring showed recovery instability, with cumulative recovery rates for WHO Grades 2–4 at 82.1%, 50.8%, and 28.2%, respectively, while peaks at 50.8%, 28.3%, and 7.3%.
Conclusions
Brain tumor-specific symptoms significantly impact general work ability recovery. Recovery instability was observed across all patients, underscoring the importance of targeted symptom management, personalized care, and sustained follow-up to improve quality of life.
Keywords: glioma, risk factors, work
Key Points.
Brain tumor-specific symptoms significantly impact general work ability recovery compared to general symptoms.
Recovery instability highlights the need for long-term follow-up and personalized care.
Importance of the Study.
Glioma research has increasingly emphasized quality of life alongside traditional survival metrics, emphasizing functional outcomes, symptom burden, and social reintegration, including the ability to work. Unlike previous studies that emphasized return-to-work rates, this study evaluates general work ability as a broader and more inclusive measure of work capacity. Our predictive models, with pre- and post-surgery AUCs of 0.78 and 0.82, respectively, provide robust tools for assessing recovery potential. Additionally, we demonstrate that brain tumor-specific symptoms have a stronger negative impact on recovery than general symptoms, emphasizing the need for targeted symptom management tailored to patients with glioma. Long-term monitoring reveals recovery instability across different tumor grades, underscoring the importance of sustained follow-up and personalized care plans. Our research provides actionable insights for clinical practice, highlights the necessity of integrating work ability assessments into long-term care strategies, and contributes to enhancing the quality of life for patients with glioma.
Introduction
Gliomas are the most common histological type of primary central nervous system cancer, and accounted for 26.3% of all CNS tumors.1–3 Adolescent and young adult (AYA, aged 15-39 years) patients experience a considerable disease burden from gliomas, which are a significant contributor to cancer-related deaths in this age group, therefore exerting a substantial influence on the working-age demographic.1 Recent glioma research has increasingly incorporated quality of life4–7 considerations alongside traditional survival metrics such as overall survival (OS) and progression-free survival (PFS). This shift has led to greater emphasis on functional outcomes, symptom burden, and social reintegration as key aspects of patient-centered care. The World Health Organization’s International Classification of Functioning, Disability and Health (ICF) provides a comprehensive framework for guiding rehabilitation, encompassing three primary dimensions: impairment, activity limitation, and participation restrictions.8
Within this ICF framework, work ability is a key component of social participation, influencing economic stability and personal identity.9 Cancer survivors perceive the recovery of work ability as an imperative means of regaining control of their lives, and it is also associated with better quality of life.9–11 However, 30-50% of survivors experience significant cognitive and functional impairment, posing considerable challenges to the restoration of their work ability. Moreover, even mild cognitive changes can profoundly impact on higher-level functional goals, underscoring the importance of investigating work ability as an independent outcome.6
Prior studies on the work ability of patients with glioma have focused on the ability to return to work, which is commonly evaluated as the proportion of patients who resume employment at specific time points12–15 but is less strongly associated with work capacity. The return to work is a practical and objective metric; however, it is influenced by factors beyond the actual work capacity, such as socioeconomic factors such as social security; for example, those with impaired work ability but act as the sole breadwinner in their families often feel compelled to get back to work under economic pressure.12,16 Moreover, there is limited research on the long-term sustainability of work ability in patients with glioma, leaving gaps in our understanding of their work trajectories.17,18
To address these limitations, our study focused on the “general work ability,” which is based on the most common occupation in the labor market that only requires minimal training,19 rather than the patients’ preoperative or long-term occupation. This approach helps to better capture work interference as a functional limitation, independent of specific occupational and socioeconomic influences.
We developed preoperative and postoperative logistic regression models to predict general work ability 6 months post-surgery and used Cox regression analysis to identify long-term risk factors. Additionally, we assessed the relative impact of brain tumor-specific symptoms (e.g. cognitive deficits, physical weakness) and general symptoms (e.g. fatigue, pain) on work ability, providing insights into the need for targeted symptom management. Finally, we conducted long-term monitoring to evaluate the stability of postoperative work ability across WHO grades, contributing to personalized care and long-term career planning for patients with glioma.
Methods
Study Design and Participants
A retrospective cohort study was carried out at Huashan Hospital, Fudan University, Shanghai, China, from March 2010 to December 2018. Standard treatment paradigms for patients with WHO Grade 2–4 gliomas at Huashan Hospital are detailed in Supplementary Note S1. After applying exclusion criteria (Fig. 1), 1847 patients met the inclusion criteria: age 18–64, WHO Grade 2–4 glioma (2021 CNS WHO criteria), and first tumor resection at Huashan Hospital.
Figure 1.
Flow diagram of patient selection. (A) The data were collected from the Department of Neurosurgery of Huashan Hospital, Fudan University, Shanghai, China. Patients aged 18–64 with WHO grade 2–4 were recruited in the basic database, as elder are defined as aged 65 and above according to WHO definition. Those with 6-month outcome data were incorporated into the Logistic dataset, and Logistic prediction models were established for both pre- and postoperative, and the follow-up data were used to construct Cox proportional hazards models. (B) Patients had first surgery in other hospitals were excluded for lack of follow-up data.
Then, from the basic dataset, we developed Cox and Logistic Datasets:
Cox Dataset:
Included patients with ≥ 1 valid post-surgery follow-up and baseline data on demographics (age, sex), clinical status (Preoperative KPS score), IDH mutation status, Ki-67 proliferation index, tumor location, and treatment (radiotherapy, chemotherapy).
Logistic Dataset:Logistic Dataset:
Included patients with 6-month general work ability status which must meet one of the following: follow-up record at exactly 6 months post-surgery with complete M.D. Anderson Symptom Inventory-Brain Tumor Module (MDASI-BT) data, regained general work ability or verified death within 6 months. Preoperative analysis focused on demographics, KPS, tumor location, and chief complaints; while postoperative analysis further incorporated treatment, WHO grade, IDH mutation status and Ki-67 proliferation index.
Definition and Measurement of General Work Ability
General work ability wass defined as the capacity to perform work accessible to most people with minimal training, considering the physical and mental health needed, including housework.19 Minimal training refers to tasks that can be learned within a short period without requiring specialized education or extensive vocational preparation.20
This was assessed using the Chinese version of the M.D. Anderson Symptom Inventory-Brain Tumor Module (MDASI-BT), a validated tool measuring both symptom severity and interference with daily life activities.21 The Chinese version has been validated for reliability and applicability in Chinese-speaking patients with glioma.22 The items and scoring of the MDASI-BT are detailed in Supplementary Note S2.
Frequency and timing of MDASI-BT assessments:
Patients were followed up every three months postoperatively (months 3, 6, 9, and 12), except for the first 2 at months 1 and 3. Beyond 12 months, follow-ups were conducted at longer intervals to monitor long-term recovery. MDASI-BT completion was mandatory for valid follow-ups. Some patients missed follow-ups due to mortality or changes in follow-up willingness.
Symptom assessment:
Part I of the MDASI-BT assesses the severity of symptoms experienced in the past 24 hours, using an 11-point scale ranging from 0 (not present) to 10 (as severe as imaginable). The scale evaluates 13 core symptoms, including pain, fatigue, and nausea, as well as 9 brain tumor-specific symptoms, such as weakness on one side of the body, difficulty speaking, and seizures.
General work ability assessment:
General work ability was assessed using the item How much have your symptoms interfered with your work (including housework) in the last 24 hours? was used to measure general work ability. This was measured on an 11-point scale, ranging from 0 (no interference) to 10 (complete interference). Having a defined numeric cutoff of continuous variables is clinically more useful, as it can facilitate interpretation for clinicians.23
Based on prior literature, a work ability interference score ≤ 1 was categorized as low-interference status, indicating no or mild interference, while a score > 1 was categorized as high-interference status.24 For analyses conducted at single time points, work ability status was classified accordingly (applied in the following work ability trajectory analysis).
Definition of work ability recovery:
We defined recovery as the first achievement of low-interference status (work ability interference score ≤ 1) after surgery, cause all patients following surgery were initially classified as having high-interference status due to the acute postoperative condition. This definition was applied in both Logistic and Cox analysis.
Study Groups
Patients were classified into 2 groups for Logistic analysis:
Recovered group: Patients achieving at least one follow-up score ≤ 1 within 6 months.
Unrecovered group: Patients who did not recover work ability or died within 6 months.
Statistical Analysis
Descriptive statistics summarized the whole logistic dataset, with continuous variables presented as the mean ± standard deviation or median [interquartile range], and analyzed using Student’s t-tests or rank-sum tests. Categorical variables were analyzed using chi-square tests or continuity-corrected chi-square tests.
Logistic regression models:
Preoperative and postoperative logistic regression models were developed to predict general work ability recovery status within 6 months, incorporating core and brain tumor-specific symptoms as independent variables. SHAP values and 10-fold cross-validation assessed feature importance and model robustness. ROC curves and AUC evaluated predictive performance, with internal validation via bootstrapping (n = 1000).
Cox proportional hazards models:
Cox proportional hazards models for each WHO Grade 2-4 were used to identify the risk factors influencing the recovery of general work ability.
Kaplan–Meier analysis:
Kaplan‒Meier survival analysis estimated cumulative general work ability recovery rates over time, with differences in recovery rates among grades assessed using the log-rank test.
General work ability status trajectory:
General work ability status trajectories were depicted over time using stacked bar charts and line graphs based on follow-up data at each time point.
The threshold was used for statistical significance assessments. Data analyses were conducted using R software (version 4.3.2) and the Storm Statistical Platform (www.medsta.cn/software).
Sensitivity Analysis
We also conducted a sensitivity analysis of the logistic regression and Cox proportional hazards models using different cutoff values to demonstrate that the results remain relatively consistent across similar cutoffs, indicating that the findings are not particularly sensitive to any specific cutoff.
Ethics approval
The research protocol was approved by the Ethical Review Committee of Huashan Hospital, Fudan University (No.KY2015-256). Written informed consent was obtained from all participants in the Central Tissue Bank for Neurological Disorders at Huashan hospital.
Results
Patient Characteristics
Among the 342 patients with glioma, 65.2% (n = 223) regained general work ability within 6 months, while 34.8% (n = 119) did not. The recovered group was significantly younger (42.25 ± 9.59 vs. 49.39 ± 10.44 years, p < 0.001) and had a higher proportion of WHO Grade 2 gliomas (151/223 [67.71%] vs. 34/119 [28.57%], p < 0.001), better preoperative KPS scores (87.24 ± 8.90 vs. 81.53 ± 14.60, p < 0.001), and lower Ki-67 indices (4.00 [2.00–8.00] vs. 10.00 [5.00–19.00], p < 0.001). IDH mutations were more common in the recovered group (122/176 [69.32%] vs. 29/87 [33.33%], p < 0.001). Radiotherapy was more frequently received in the recovered group (147/220 [66.82%] vs. 60/119 [50.42%], p = 0.003), while chemotherapy did not differ significantly (122/220 [55.45%] vs. 56/119 [47.06%], p = 0.140).
Postoperative complications were more frequent in the unrecovered group, particularly difficulty remembering (11/102 [10.78%] vs. 0/223 [0.00%], p < 0.001), weakness on one side (25/102 [24.51%] vs. 3/223 [1.35%], p < 0.001), and difficulty speaking (20/102 [19.61%] vs. 2/223 [0.90%], p < 0.001). The postoperative complications were not consistently documented in medical records; instead, symptom burden was assessed using the MDASI-BT, aligning specific symptoms with known complications to provide a patient-centered evaluation. Detailed correspondence is provided in Supplementary Note S3. Postoperative KPS scores were also lower (82.51 ± 20.27 vs. 94.96 ± 6.83, p < 0.001) in the unrecovered group.
Detailed patient selection and group characteristics are shown in Figure 1 and Table 1.
Table 1.
Demographic and clinical data of patients in the general work ability recovered and unrecovered groups.
| Variable (N available) |
Work ability status (N = 342) | t/z/χ2 | P-valueb | |
|---|---|---|---|---|
| Unrecovered (N = 119, 34.80%)a |
Recovered (N = 223, 65.20%)a |
|||
| Sex male No. (%) (342) | 71 (59.66) | 136 (60.99) | χ2 = 0.057 | .812 |
| Age (years at diagnosis) (342) | 49.39 ± 10.44 | 42.49 ± 9.59 | t = 6.151 | <.001*** |
| WHO Gradec (342) WHO Grade 2 |
34 (28.57) |
151 (67.71) |
χ2 = 69.490 |
<.001*** |
| WHO Grade 3 | 19 (15.97) | 41 (18.39) | ||
| WHO Grade 4 Classification of tumorsd (213) IDH mutant, astrocytoma IDH mutant, oligodendroglioma IDH-wildtype, glioblastoma |
66 (55.46) 19 (26.39) 6 (8.33) 47 (65.28) |
31 (13.90) 86 (60.99) 35 (24.82) 20 (14.18) |
χ2=57.865 | <.001*** |
| Preoperative KPS (342) | 81.53 ± 14.60 | 87.24 ± 8.90 | t = -3.885 | <.001*** |
| Ki-67 (Q₁-Q₃) (328) | 10.00 (5.00-19.00) | 4.00 (2.00-8.00) | z = -5.843 | <.001*** |
| IDH-1/IDH-2 + No. (%) (263) MGMT + no. (%)e (225) |
29 (33.33) 46 (53.49) |
122 (69.32) 57 (41.01) |
χ2 = 30.834 χ2=3.334 |
<.001*** .068 |
| Radiotherapy no. (%) (339) | 60 (50.42) | 147 (66.82) | χ2 = 8.734 | .003** |
| Chemotherapy no. (%) (339) Extent of resection no. (%)f (331) GTR No. (%) STR (residual < 10%) No. (%) Lobectomy Partial Resection |
56 (47.06) 107 (93.86) 6 (5.26) 1 (0.88) 0 (0.00) |
122 (55.45) 206 (94.93) 9 (4.15) 1 (0.46) 1 (0.46) |
χ2 = 2.183 |
.140 .874 |
| Chief complaints (339) | ||||
| Memory deterioration no. (%) | 9 (7.76) | 6 (2.46) | χ2 = 4.635 | .031* |
| Speech disorder no. (%) | 5 (4.31) | 6 (2.69) | χ2 = 0.226 | .634 |
| Muscle weakness no. (%) | 9 (7.76) | 14 (6.28) | χ2 = 0.264 | .607 |
| Impaired consciousness no. (%) | 25 (21.55) | 68 (30.49) | χ2 = 3.065 | .080 |
| Seizure no. (%) | 36 (31.03) | 91 (40.81) | χ2 = 3.110 | .078 |
| Vomiting no. (%) | 12 (10.34) | 19 (8.52) | χ2 = 0.306 | .580 |
| Nausea no. (%) | 8 (6.90) | 7 (3.14) | χ2 = 2.548 | .110 |
| Dizziness no. (%) | 15 (12.93) | 24 (10.76) | χ2 = 0.353 | .553 |
| Headache no. (%) | 50 (43.10) | 72 (32.29) | χ2 = 3.875 | .049* |
| Tumor location (328) | ||||
| Insular lobe no. (%) | 9 (8.18) | 13 (5.96) | χ2 = 0.575 | .448 |
| Corpus callosum no. (%) | 9 (8.18) | 8 (3.67) | χ2 = 3.029 | .082 |
| Parietal lobe no. (%) | 13 (11.82) | 22 (10.09) | χ2 = 0.229 | .633 |
| Occipital lobe no. (%) | 6 (5.45) | 9 (4.13) | χ2 = 0.295 | .587 |
| Temporal lobe no. (%) | 39 (35.45) | 68 (31.19) | χ2 = 0.604 | .437 |
| Frontal lobe no. (%) Post-surgery complicationg(325) Difficulty remembering no. (%) Weakness on one-side no. (%) Difficulty understanding no. (%) Difficulty speaking no. (%) Seizure no. (%) Postoperative average KPS (324) |
64 (58.18) 11 (10.78) 25 (24.51) 16 (15.69) 20 (19.61) 3 (2.94) 82.51 ± 20.27 |
138 (63.30) 0 (0.00) 3 (1.35) 0 (0.00) 2 (0.90) 1 (0.45) 94.96 ± 6.83 |
χ2 = 0.810 χ2 = 21.70 χ2 = 47.70 χ2 = 36.79 χ2 = 38.83 χ2 = 1.81 t = -6.05 |
.368 <.001*** <.001*** <.001*** <.001*** .179 <.001*** |
aThe unrecovered group (N = 119) consisted of patients who did not regain general work ability within 6 months post-surgery, including those who died within this period. The recovered group (n = 223) included patients who regained general work ability within 6 months post-surgery, without considering any subsequent changes in their general work ability status.
b*: P < .05, **: P < .01, ***: P < .001.
cIn our cohort, the proportion of Grade 4 patients is relatively low due to the less comprehensive baseline information compared to those with Grade 2.
dSome patients lacked the WHO 2021 tumor classification information. Due to high missing rates (37.7%), this variable was not included in subsequent analyses.
eThe missing rate of the MGMT was also high, and the tumor volume data were unavailable. Therefore, these variables are also not presented in following analysis.
fThe extent of resection was assessed using postoperative MRI within 3 days after surgery.
gPostoperative complications were not consistently documented in medical records; instead, symptom burden was assessed using the MDASI-BT, aligning specific symptoms with known complications to provide a patient-centered evaluation.
Predictive Analysis and Symptom Evaluation
The preoperative predictive model (shown in Supplementary Table S4) for 6-month general work ability recovery yielded an AUC of 0.73, while the postoperative model (shown in Supplementary Table S5) improved to an AUC of 0.82, indicating strong predictive accuracy (Figure 2, Panels A and B). Both models identified age, KPS scores, and tumor grade as key predictors. WHO Grade 3 and Grade 4 gliomas were associated with significantly reduced odds of recovery compared to Grade 2 (OR 0.38 [0.14-0.99] and 0.11 [0.04-0.34], respectively, p < 0.05). Radiotherapy emerged as a positive predictor in the postoperative model (OR 2.77 95% CI 1.38-5.57, p = 0.004). Internal validation using bootstrap resampling confirmed the stability of these models.
Figure 2.
Predictive models and the importance of symptom-related factors for general work ability recovery. Panels A and B show the ROC curves of preoperative and postoperative predictive models for 6-month general work ability recovery, including alternative models based on other combinations of variables. Panel C shows the SHAP summary plot, illustrating the association strength between selected variables and the likelihood of general work ability recovery. The width of the horizontal bars represents the magnitude of each variable’s impact on the association, the color gradient indicates the magnitude of variable values, and the x-axis orientation shows the association probability with higher (right) or lower (left) general work ability recovery. Abbreviations: SHAP = SHapley Additive exPlanations.
The SHAP visualization plot, derived from a logistic regression model with 10-fold cross-validation (Figure 2. C), demonstrated the differential impact of various factors on the recovery of general work ability at 6 months. Symptoms specific to brain tumors adversely affected the recovery of general work ability over this period. Notably, although core symptoms were the third most impactful factor, the analysis revealed no discernible correlation between the severity of core symptoms and the recovery of general work ability at 6 months.
Cox Proportional Hazards Modeling and Kaplan–Meier Survival Analysis
Cox regression analyses for each WHO Grade 2-4 provided insight into the factors influencing the recovery of general work ability in postoperative patients with glioma (Supplementary Tables S6, S7, S8).
For Grade 2 patients with glioma, male sex (HR 1.55, 95% CI 1.26-1.90, p < 0.001) and younger age (HR 0.95, 95% CI 0.94-0.96, p < 0.001) were associated with likelihood of general work ability recovery in the multivariate analysis. Frontal lobe tumor location was also a significant factor (HR 1.36, 95% CI 1.09-1.69; p = 0.007). In contrast to the logistic regression results, radiotherapy served as a negative factor (HR 0.71, 95% CI 0.58-0.88; p = 0.002).
For Grade 3 patients, male sex was a favorable factor (HR 1.80, 95% CI 1.20-2.70, p = 0.005), and age continued to be a significant risk factor (HR 0.96, 95% CI 0.94-0.98, p < 0.001). Radiotherapy had a negative effect on the univariate analysis (HR 0.58, 95% CI 0.34-0.99; p = 0.048) but was not significant in the multivariate analysis (HR 0.68, 95% CI 0.40-1.17; p = 0.163).
In the case of Grade 4 patients, IDH mutation status was significant in the univariable analysis (HR 2.00, 95% CI 1.14-3.50, p = 0.016), but not in the multivariable analysis. Age remained a consistent risk factor (HR 0.96, 95% CI 0.94-0.98, p < 0.001), while sex did not impact on general work ability recovery for this group.
Kaplan–Meier survival curves (shown in Figure 3) showed cumulative recovery rates of 82.1% (Number at risk, Grade 2: 652 at baseline, 158 at 25 months, 56 at 50 months, 12 at 75 months), 50.8% (Grade 3: 275 at baseline, 76 at 25 months, 24 at 50 months, 1 at 75 months), and 28.2% (Grade 4: 444 at baseline, 60 at 25 months, 14 at 50 months, 6 at 75 months) (p < 0.001), reflecting the progressive impact of tumor grade on recovery and the decreasing number at risk over time due to censoring and disease progression. And the median time for cumulative recovery was 18 months for WHO Grade 2 and 48 months for WHO Grade 3. For WHO Grade 4, the median recovery time was not available due to insufficient recovered cases.
Figure 3.
Cumulative general work ability recovery analysis for postoperative patients with WHO Grades 2–4 gliomas. The figure shows the cumulative general work ability recovery rates over time for patients with glioma. The final cumulative rates of grades 2, 3, and 4 are 0.821, 0.508 and 0.282, respectively. The recovery rates decreased with increasing grade (P < .001), indicating a decreased likelihood of recovery in more advanced tumor grades.
Further Investigation into the Long-Term Effects of Radiotherapy on General Work Ability
Logistic regression identified radiotherapy as a positive predictor of general work ability recovery within 6 months, while Cox models indicated a negative long-term association, particularly in Grade 2 patients. To clarify whether the observed discrepancy stems from delayed adverse effects of radiotherapy or other tumor aggressiveness factors associated with the decision to undergo radiotherapy, or disease progression, we conducted further analysis. To minimize the influence of tumor aggressiveness and treatment decision, we performed a stratified analysis by tumor grade (WHO Grades 2, 3, and 4) and further stratified Grade 2 gliomas by IDH mutation status.
We applied logistic regression, with the dependent variable being general work ability deterioration, defined as the transition from low-interference status (work interference score ≤ 1) to high-interference status (score > 1). The control group included patients who achieved and maintained low-interference status throughout follow-up. Independent variables were drawn from the original postoperative logistic model, with tumor recurrence added to assess its association with work ability decline.
Due to complete separation in logistic models for Grades 3, 4, and IDH-wildtype Grade 2 patients (all recurrent cases experienced work ability deterioration, detailed in Supplementary Tables S9,S10), we applied Firth logistic regression, which confirmed tumor recurrence as the strongest factor across all subgroups (OR > 10 in Grade 2 patients and OR > 180 in higher-grade patients, the complete separation made OR extremely high, detailed in Supplementary Tables S11–S14). While radiotherapy was included in the models, its effect was not statistically significant in those subgroups after adjusting for confounders. This suggests that the previously observed association between radiotherapy and work ability deterioration is largely explained by its correlation with tumor aggressiveness, rather than a direct causal effect.
General Work Ability Status Trajectory
The trajectories of general work ability status post-surgery for patients with WHO Grades 2-4 are presented in Figure 4 and demonstrate significant variability over time. The line graphs illustrate the low-interference status proportion, defined as the proportion of patients who were in low-interference status out of the total population at each time point, which calculated as Low-Interference Proportion = Low-Interference Status / (Low-Interference Status + High-Interference Status + Deceased). Each WHO grade showed distinct patterns in the trajectories.
Figure 4.
General work ability status trajectory post-surgery by WHO grade. The figure presents the trajectory of general work ability status over time for patients with WHO Grades 2–4 glioma in Panels A, B, and C, respectively. The stacked bar charts illustrate the distribution of patient statuses at different time points. The line graphs within each panel represent the low-interference proportion, calculated as Low-Interference Status / (Low-Interference Status + High-Interference Status + Deceased). The highest observed low-interference proportion was 50.8% for Grade 2 patients at 30 months, 28.3% for Grade 3 patients at 30 months, and 7.6% for Grade 4 patients at 12 months.
For patients with Grade 2 gliomas, the maximum low-interference proportion was 50.8% (213/[213 + 180 + 26]), which was observed at 30 months post-surgery. Patients with Grade 3 gliomas reached a maximum low-interference proportion of 28.3% (52/[52 + 75 + 57]) at the same follow-up time. In contrast, patients with grade 4 gliomas exhibited a much lower maximum low-interference proportion of 7.6% (28/[28 + 239 + 103]) at 12 months. These findings indicate that higher-grade patients with glioma have a significantly diminished likelihood of regaining work ability, with the rate decreasing sharply as the tumor grade increases.
Sensitivity Analysis
Our sensitivity analysis explores the impact of employing alternative cutoff values of 1 and 2 in our models. For the logistic models, AUCs for cutoff 1 are 0.73 preoperatively and 0.83 postoperatively; for cutoff 2, AUCs are 0.72 preoperatively and 0.83 postoperatively. And the AUCs for pre- and postoperative logistic models under both cutoffs are consistent, indicating stable predictive performance. Meanwhile, the significant factors identified in these models, as well as the risk factors in Cox models stratified by WHO grade, show substantial overlap between the 2 cutoffs. These findings suggest that our initial cutoff selection is justified and that our study’s outcomes are not overly sensitive to this parameter. The results are displayed in Supplementary Tables S15 and 16.
Discussion
Regaining work ability is a positive step toward rehabilitation, as it impacts family finances and serves as a symbol of recovering personal identity and a sense of normalcy.25–27
In this study, we investigated general work ability, a subjective self-assessment of work ability, defined based on the most common types of work in the labor market that only requires a short period of training to perform, including housework,19 to better capture work interference as a functional limitation independent of specific occupational and socioeconomic influences. We constructed logistic predictive models for the pre- and postoperative periods to assess the 6-month recovery of general work ability and Cox models to investigate associated risk factors. Additionally, brain tumor-specific symptoms, such as cognitive deficits and physical weakness, were found to have a significant negative impact on work ability recovery, surpassing the influence of general symptoms like fatigue and pain. We also compared the differences between cumulative general work ability recovery and the trajectory of general work ability status. It was observed that even when patients with glioma recovered their general work ability postoperatively, this status was not stable. This underscores the necessity for long-term monitoring of general work ability in patients with glioma and the implementation of specific care interventions.
According to the logistic predictive models, age, preoperative KPS, and WHO grade were found to be important predictors for work ability. Notably, radiotherapy served as a positive indicator of general work ability recovery 6 months post-surgery in our study, but had negative effect in the long-term Cox models (particularly in Grade 2 patients). To resolve this discrepancy, we conducted stratified analyses to determine whether this observed association was due to delayed adverse effects of radiotherapy, or if it was confounded by tumor aggressiveness and recurrence. We stratified by tumor grade and further by IDH status in Grade 2 patients. Given the issue of complete separation (where all Grade 3 and 4 patients who recurred experienced work ability deterioration), we applied Firth logistic regression to obtain more stable estimates. Across all subgroups, tumor recurrence consistently showed the strongest association with work ability deterioration, with extremely high odds ratios even after correction for separation. In contrast, radiotherapy was not independently associated with work ability deterioration after adjusting for recurrence, suggesting that its observed effect was largely confounded by tumor aggressiveness rather than representing a direct causal relationship.
Kaplan–Meier curves (Figure 3) show optimistic cumulative recovery rates for WHO Grades 2–4 gliomas (82.1%, 50.8%, and 28.2%, respectively). However, dynamic work ability status trajectories (Figure 4) reveal lower peak values of low-interference status proportion (50.8%, 28.3%, and 7.6%, respectively), emphasizing that initial recovery may not sustain due to tumor recurrence or mortality. This observation underscores the critical importance of viewing work ability as a fluctuating state rather than a definitive outcome, which calls for a longitudinal perspective in monitoring general work ability.28,29 Understanding the sustainability of work ability recovery is essential, as transient work ability holds limited value for a patients’ long-term quality of life. Moreover, continuous monitoring of general work ability allows for adjustments of rehabilitation trials based on patient’s status.
Symptom management is an essential part of health care for cancer patients.30 The MDASI-BT has the benefits of brevity, the ability to measure multiple symptoms, and patient ratings of the interference of symptoms on activities of daily living as a measure of the burden of symptoms.31 We specifically utilized the Chinese version of the MDASI-BT to explore the impact of core symptoms and symptoms specific to brain tumors on patients’ postoperative self-assessed general work ability status. Our findings indicate that symptoms specific to brain tumors significantly influence patients’ general work ability status, underscoring the critical need for targeted interventions in managing patients with glioma’ work ability recovery beyond the general approaches applied to all cancer patients.
In contrast to our study, previous research often utilizes return to work to assess work ability recovery, which is impacted by a variety of subjective and objective factors beyond their real work ability that significantly influence employment outcomes.32,33 The financial role in family and social security may impact patients’ willingness to return to work. For instance, those with impaired work ability but who act as the sole breadwinner in their families often feel compelled to return to work under economic pressure.14 Additionally, patients’ previous occupation and career expectations can also affect their occupation decisions. Professionals working in extremely specialized domains, such as engineers or pilots, may face significant obstacles when trying to return to their prior positions due to the high degree of accuracy needed.12 Manually demanding jobs were also negatively correlated with return to work.13 Moreover, patients’ work expectations, might influence their job decisions, such as whether they resume their previous role or opt for full-time employment.34 Incorporating both metrics in future studies could provide a holistic understanding of postoperative work capacity and inform tailored care strategies.
Our study has several limitations. As this was a single-center study, it underwent internal validation only, potentially introducing bias. Additionally, the extended follow-up period resulted in more lost-to-follow-up participants, which could affect the accuracy of the observed trajectory in general work ability status. Furthermore, while our research emphasizes the concept of self-assessed general work ability, we acknowledge that the inclusion of housework as part of the work ability assessment may not fully reflect the physical and cognitive demands of unskilled labor in the job market. Housework, although a relevant measure of daily functioning, may not be directly comparable to the intensity of tasks involved in unskilled labor.
Despite this, our focus on self-assessed general work ability provides a valuable measure of patients’ functional recovery, offering a standardized and symptom-driven assessment that is independent of occupational and socioeconomic variability. Utilizing the widely adopted MDASI-BT, the general work ability is particularly suited for comparing postoperative work ability across different studies or patient groups. It offers a universal application that transcends individual patient backgrounds. Further research could combine these 2 metrics to provide a more holistic understanding of the factors influencing postoperative work ability in patients with glioma, thereby contributing to the development of tailored care strategies.
Supplementary Material
Acknowledgments
The authors thank Dr. Zhenxiao Wang for her work on postoperative follow-up data collection.
Contributor Information
Xin’er Yuan, Department of Biostatistics, School of Public Health & National Clinical Research Center for Aging and Medicine, Huashan Hospital, Fudan University, Shanghai, China.
Jie Zhang, Neurosurgical Institute of Fudan University, Shanghai, China; Shanghai Key Laboratory of Brain Function and Restoration and Neural Regeneration, Shanghai, China; National Center for Neurological Disorders, Shanghai\, China; Department of Neurosurgery, Huashan Hospital, Shanghai Medical College, Fudan University, Shanghai, China.
Shuai Wu, Neurosurgical Institute of Fudan University, Shanghai, China; Shanghai Key Laboratory of Brain Function and Restoration and Neural Regeneration, Shanghai, China; National Center for Neurological Disorders, Shanghai\, China; Department of Neurosurgery, Huashan Hospital, Shanghai Medical College, Fudan University, Shanghai, China.
Chen Luo, Neurosurgical Institute of Fudan University, Shanghai, China; Shanghai Key Laboratory of Brain Function and Restoration and Neural Regeneration, Shanghai, China; National Center for Neurological Disorders, Shanghai\, China; Department of Neurosurgery, Huashan Hospital, Shanghai Medical College, Fudan University, Shanghai, China.
Ye Yao, Department of Biostatistics, School of Public Health & National Clinical Research Center for Aging and Medicine, Huashan Hospital, Fudan University, Shanghai, China.
Funding
This work is supported by the STI 2030—Major Projects No. 2022ZD0209900, the Innovation Program of Shanghai Municipal Education Commission (2023ZKZD13), and the National Natural Science Foundation of China (No.82327807).
Conflict of interest statement. The authors have no competing interests to declare that are relevant to the content of this article.
Consent to participate
Informed consent was obtained from all individual participants included in the study.
Data, material and/or code availability
Data supporting the results of this study are available from the corresponding author (YY) upon reasonable request and approval.
Author contributions
Concept and study design: Xin’er Yuan, Jie Zhang, Ye Yao; Acquisition of data: Jie Zhang, Chen Luo, Shuai Wu; Statistical analysis: Xin’er Yuan; Interpretation of results: Xin’er Yuan, Jie Zhang; Drafting of the manuscript: Xin’er Yuan, Jie Zhang; Final approval of the submitted version: Jie Zhang, Ye Yao.
References
- 1. Weiser A, Sanchez Bergman A, Machaalani C, et al. Bridging the age gap: a review of molecularly informed treatments for glioma in adolescents and young adults. Front Oncol. 2023;13:1254645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Patel AP, Fisher JL, Nichols E, et al. Global, regional, and national burden of brain and other CNS cancer, 1990–2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 2019;18(4):376–393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Ostrom QT, Price M, Neff C, et al. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2016—2020. Neuro-Oncology. 2023;25(Suppl_4):iv1–iv99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Di Perri D, Jmil S, Lawson TM, et al. Health-related quality of life and cognitive failures in patients with lower-grade gliomas treated with radiotherapy. Cancer Radiother. 2023;27(3):219–224. [DOI] [PubMed] [Google Scholar]
- 5. Rimmer B, Bolnykh I, Dutton L, et al. Health-related quality of life in adults with low-grade gliomas: a systematic review. Qual Life Res. 2023;32(3):625–651. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Van Dyk K, Wall L, Heimberg BF, et al. Daily functioning in glioma survivors: associations with cognitive function, psychological factors and quality of life. CNS Oncol. 2022;11(02):CNS84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Aaronson NK, Taphoorn MJB, Heimans JJ, et al. Compromised health-related quality of life in patients with low-grade glioma. J Clin Oncol. 2011;29(33):4430–4435. [DOI] [PubMed] [Google Scholar]
- 8. Stucki G, Cieza A, Melvin J.. The international classification of functioning, disability and health (ICF): A unifying model for the conceptual description of the rehabilitation strategy. J Rehabil Med. 2007;39(4):279–285. [DOI] [PubMed] [Google Scholar]
- 9. Rasmussen DM, Elverdam B.. The meaning of work and working life after cancer: an interview study. Psychooncology. 2008;17(12):1232–1238. [DOI] [PubMed] [Google Scholar]
- 10. de Boer AG, Torp S, Popa A, et al. Long-term work retention after treatment for cancer: a systematic review and meta-analysis. J Cancer Surviv. 2020;14(2):135–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Di Meglio A, Menvielle G, Dumas A, et al. Body weight and return to work among survivors of early-stage breast cancer. ESMO Open. 2020;5(6):e000908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Pascual JSG, Duffau H.. The need to consider return to work as a main outcome in patients undergoing surgery for diffuse low-grade glioma: a systematic review. Acta Neurochir (Wien). 2022;164(10):2789–2809. [DOI] [PubMed] [Google Scholar]
- 13. Starnoni D, Berthiller J, Idriceanu TM, et al. Returning to work after multimodal treatment in glioblastoma patients. Neurosurg Focus. 2018;44(6):E17. [DOI] [PubMed] [Google Scholar]
- 14. Yoshida A, Motomura K, Natsume A, et al. Preoperative predictive factors affecting return to work in patients with gliomas undergoing awake brain mapping. J Neurooncol. 2020;146(1):195–205. [DOI] [PubMed] [Google Scholar]
- 15. Mandonnet E, De Witt Hamer P, Poisson I, et al. Initial experience using awake surgery for glioma. Neurosurgery. 2015;76(4):382–9; discussion 389. [DOI] [PubMed] [Google Scholar]
- 16. Ahn E, Cho J, Shin DW, et al. Impact of breast cancer diagnosis and treatment on work-related life and factors affecting them. Breast Cancer Res Treat. 2009;116(3):609–616. [DOI] [PubMed] [Google Scholar]
- 17. Duijts SFA, Kieffer JM, van Muijen P, van der Beek AJ.. Sustained employability and health-related quality of life in cancer survivors up to four years after diagnosis. Acta Oncol. 2017;56(2):174–182. [DOI] [PubMed] [Google Scholar]
- 18. Arndt V, Koch-Gallenkamp L, Bertram H, et al. Return to work after cancer. A multi-regional population-based study from Germany. Acta Oncol. 2019;58(5):811–818. [DOI] [PubMed] [Google Scholar]
- 19. Tengland PA, Tengland PA.. The concept of work ability. J Occup Rehabil. 2011;21(2):275–285. [DOI] [PubMed] [Google Scholar]
- 20. ODISP I. Social Security Ruling: SSR 1983.-: SSR 1910. [Google Scholar]
- 21. Armstrong TS, Mendoza T, Gring I, et al. Validation of the M.D. Anderson symptom inventory brain tumor module (MDASI-BT). J Neurooncol. 2006;80(1):27–35. [DOI] [PubMed] [Google Scholar]
- 22. Shuying PU, Qianqian YAO, Wei JIANG, et al. Reliability and validation study of the Chinese version of M.D. Anderson symptom inventory brain tumor module. Chin J Nurs. 2021;56(6):855–860. [Google Scholar]
- 23. Armstrong TS, Vera-Bolanos E, Gning I, et al. The impact of symptom interference using the MD Anderson Symptom Inventory-Brain Tumor Module (MDASI-BT) on prediction of recurrence in primary brain tumor patients. Cancer. 2011;117(14):3222–3228. [DOI] [PubMed] [Google Scholar]
- 24. Leeper HE, Vera E, Christ A, et al. Association of employment status with symptom burden and health-related quality of life in people living with primary CNS tumors. Neurology. 2023;100(16):e1723–e1736. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Blinder VS, Gany FM.. Impact of cancer on employment. J Clin Oncol. 2020;38(4):302–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Blinder V, Patil S, Eberle C, Griggs J, Maly RC.. Early predictors of not returning to work in low-income breast cancer survivors: a 5-year longitudinal study. Breast Cancer Res Treat. 2013;140(2):407–416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Silvaggi F, Leonardi M, Raggi A, et al. Employment and work ability of persons with brain tumors: a systematic review. Front Hum Neurosci. 2020;14:571191. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. de Boer AGEM, Verbeek JHAM, Spelten ER, et al. Work ability and return-to-work in cancer patients. Br J Cancer. 2008;98(8):1342–1347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. de Moor JS, Kent EE, McNeel TS, et al. employment outcomes among cancer survivors in the United States: implications for cancer care delivery. J Natl Cancer Inst. 2021;113(5):641–644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Armstrong TS, Cohen MZ, Eriksen LR, Hickey JV.. Symptom clusters in oncology patients and implications for symptom research in people with primary brain tumors. J Nurs Scholarsh. 2004;36(3):197–206. [DOI] [PubMed] [Google Scholar]
- 31. Armstrong TS, Gilbert MR.. Net clinical benefit: functional endpoints in brain tumor clinical trials. Curr Oncol Rep. 2007;9(1):60–65. [DOI] [PubMed] [Google Scholar]
- 32. Taskila T, Lindbohm ML, Taskila T, et al. Factors affecting cancer survivors’ employment and work ability. Acta Oncol. 2007;46(4):446–451. [DOI] [PubMed] [Google Scholar]
- 33. Elsheikh M, Bridgman E, Lavrador JP, Lammy S, Poon MTC.. Association of extent of resection and functional outcomes in diffuse low-grade glioma: systematic review & meta-analysis. J Neurooncol. 2022;160(3):717–724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. De Witt Hamer PC, Klein M, Hervey-Jumper SL, Wefel JS, Berger MS.. Functional outcomes and health-related quality of life following glioma surgery. Neurosurgery. 2021;88(4):720–732. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.