Abstract
BACKGROUND AND OBJECTIVES:
Patients with newly diagnosed lower-grade glioma (World Health Organization grade II and III) are typically of working age. However, work resumption after surgical resection is uncertain, possibly due to loss of capacity from resection of tumor-infiltrated brain regions. Therefore, we explore the association between work resumption and the resection location in addition to other patient, tumor, and treatment characteristics.
METHODS:
This retrospective cohort consisted of adults undergoing first-time resection for lower-grade glioma between 2011 and 2020 in hospitals in France, Germany, and the Netherlands. Employment was evaluated at baseline and within 1 year after surgery to determine work resumption. The association between work resumption and patient, tumor, and treatment characteristics was analyzed using logistic regression. Resection cavities were segmented from postoperative MRI scans, registered to standard brain space and related to gray nuclei, cortical networks, and white matter tracts using atlas parcellations. To identify brain regions potentially involved with work resumption, the association between work resumption and resection location was analyzed using Bayesian hurdle regression. The identified regions and characteristics were jointly analyzed in their association with work resumption using multiple logistic regression.
RESULTS:
Of 207 patients, 181 (87%) were employed at baseline. Of these employed patients, 111 (61%) had resumed work at follow-up. Male sex, younger age, and larger extent of resection were independent significant predictors of work resumption. Resection location was not associated with work resumption.
CONCLUSION:
Almost two-thirds of patients resumed work 1 year after surgery. Work resumption was associated with patient characteristics (male sex and younger age) and extent of resection, but not with resection location.
KEY WORDS: Return to work, MRI, Brain neoplasms, Lesion symptom mapping, Logistic models
ABBREVIATIONS:
- AIC
Akaike information criterion
- EoR
extent of resection
- T1c
3D T1 contrast-enhanced scan series
- WHO
World Health Organization.
Patients with lower-grade glioma have a median overall survival up to 17 years for grade II gliomas and 9 years for grade III gliomas depending on molecular subtype.1 Maintaining quality of life after treatment is therefore crucial. The ability to return to work after glioma surgery contributes to normalcy and quality of life and has been of growing interest.2
Work resumption after glioma surgery has mostly been assessed in smaller cohorts from single-center studies, reporting work resumption rates between 52% and 91% in the first year.3-10 These studies found various predictors for work resumption such as younger age, male sex, sole breadwinner status, preoperative tumor volume, extent of resection (EoR), postoperative Karnofsky performance status, verbal fluency, lexical access speed, and postoperative seizures.2,4,5,7-13 In addition, surgical strategy may also affect work resumption. Intraoperative mapping specifically has been suggested to improve functional outcome and work resumption for glioma patients.14,15 However, the relationship between resection location and work resumption in patients with glioma has not been reported.5,6,15-17 A better understanding of the risk of surgery for work resumption may facilitate patient counseling and guide surgical decisions.2
Associations between functionality and lesion locations, such as the resection location, have traditionally been mapped using voxel-wise lesion symptom mapping.18 However, voxel-wise mapping does not take spatial relationships into account.19 Considering voxels in light of their overarching affected brain region is pivotal. Mapping functionality, such as ability to resume work, to interconnected brain regions may better explain dysfunction.
In this study, we examine the association between resection location and work resumption after lower-grade glioma surgery using region-wise lesion symptom mapping to potentially identify brain regions at risk, in addition to previously established patient and tumor characteristics.
METHODS
Patients
This retrospective study included consecutive adult patients who received first-time resection for World Health Organization (WHO) 2016 grade II or III gliomas with no recurrence at 9 to 12 months follow-up, between 2011 and 2022 at 3 academic centers: Amsterdam University Medical Centers, Amsterdam, the Netherlands; Klinikum rechts der Isar of the Technical University Munich, Germany; and Hôpital Lariboisière Paris, France. Patient data were retrieved from prospective electronic databases. Ethical approval was granted by the Medical Ethical Committee of Amsterdam UMC location Vrije Universiteit Amsterdam, and written informed consent was provided by all patients. The study adheres to Strengthening the Reporting of Observational Studies in Epidemiology guidelines.20
Work Resumption as Outcome Measure
Occupational status was determined at presentation and 9-12 months after surgery. Patients were also considered employed in the case of temporary sick leave because of the recent diagnosis. Work resumption was determined for patients employed at presentation. Students were considered employed. Stay-at-home parents and retired patients were considered unemployed.
Patient, Tumor, and Treatment Characteristics
Patient characteristics such as sex, age at diagnosis, and symptoms at presentation were collected. Tumor characteristics included WHO grade, preoperative tumor volume, and hemispheric tumor location in standard brain space. Treatment characteristics included use of intraoperative stimulation mapping, EoR, and adjuvant therapies — radiotherapy and chemotherapy. EoR was calculated by percentage of tumor volume reduction after surgery in standard space. Type of work and skill level were classified using the International Standardized Classification of Occupations.21 International Standardized Classification of Occupations classifies jobs into 4 skill levels: skill level 1 involves simple and routine physical and manual tasks, skill level 2 involves operating machinery and electronic equipment, skill level 3 involves complex skills requiring knowledge in a specialized field, and skill level 4 involves complex problem solving and decision making based on extensive theoretical knowledge in a specialized field.
Resection Location
MRI scans were retrieved from PACS systems. Scans had been acquired before surgery, within 3 days after surgery and months after surgery in follow-up. At each time point, we collected a T1-weighted scan without and with gadolinium administration (T1c), a 3D fluid-attenuated inversion recovery scan and a T2-weighted scan if available.
To define resection location, postoperative resection cavities were manually segmented by IK, supervised by PW and FB. Cavities and residual tumor volumes were segmented on the 3- to 6-month follow-up scans to avoid surgical artefacts, if unavailable, earlier postoperative scans were used. Segmented volumes were outlined in Brainlab iPlan v3.0 software (Brainlab) and exported on T1c series.
Resection distribution maps were constructed as detailed elsewhere.22,23 In short, the T1c series were registered to Montreal Neurological Institute 152 standard space using nonlinear registration with ANTs diffeomorphic SyN algorithm and lesion cost-function masking.24-26 When nonlinear registration failed quality checks, affine registration was used instead. The resection cavity segmentations were subsequently resampled to standard space at 2 mm resolution. Registered segmentations were summarized over patients at each voxel to create postoperative resection distribution maps. Registrations were performed using python v3.8.0 (Python Software Foundation).
To quantify the association of resection cavity with predefined overarching brain regions, we determined the lesion load using parcellation atlases in standard space. We weighted the lesion load of resection cavities on brain regions by dividing the resection cavity volume within a brain region by the total brain region's volume. To define brain regions, we used Yeo 7-network atlas split into left- and right-sided parcels thus including 14 cortical networks, the Harvard-Oxford atlas including 14 deep gray nuclei, and XTRACT white matter probabilistic atlases including 41 white matter tracts using a tract probability of >0%.27-32
Statistics
Differences in characteristics between centers were assessed using Kruskal-Wallis tests for continuous variables and χ2 tests for categorical variables using R v 4.1.0 (the R Foundation). Significance level was set at P-value ≤.01.
Patient, tumor, and treatment characteristics and lesion load were related to work resumption using stepwise univariable and multivariable logistic regression. Continuous variables were z-score normalized by subtracting the mean and dividing by the SD. Variables were deemed significantly associated with work resumption for P-value ≤.01. The Akaike information criterion (AIC) was used to assess model fit.
Considering the difference in covariate nature and distribution, 2 independent feature selection steps were performed. As first feature selection step, univariable logistic regression was performed for all characteristics, except lesion loads, with work resumption as dependent variable. Variables with P-value <.1 were selected for the multivariable model.
Second, a Bayesian hurdle model was used to assess differences in lesion load distributions between patients who resumed work and those who did not. The hurdle model considered a Bernoulli probability mass function to determine whether a region contained lesion, followed by an exponential distribution to estimate the lesion load in that region. Per region, the difference in posterior density distribution between patients who resumed work and those who did not was assessed as defined in the Supplemental Digital Content 1 (http://links.lww.com/NS9/A39). Methods. The top 5 resected brain regions with potential importance for work resumption were selected and used simultaneously as additional input features for the multivariable model. The Bayesian model had hierarchical dependencies with vaguely informative priors.33 The Markov chain Monte Carlo sampler used 4 chains with 1000 draws. The model was set up in Stan and run through CmdStanPy v2.33.1 in python v3.8.0 (Python Software Foundation).34
RESULTS
Work Resumption
Of 207 included patients, 181 patients (87.4%) were employed at presentation. No patients lost employment because of the onset of symptoms. There was a lower employment rate in Paris (76.3%) compared with Amsterdam (86.2%) and Munich (94.7%) though not significant (P = .018) as is shown in Figure 1. There was no differences in type of work or skill level between centers (Table 1). Of employed patients, 111 resumed work 9 to 12 months after surgery (61.3%). Work resumption data were missing for 11 patients (6.0%). Two patients were unemployed at baseline and employed at follow-up. One of these patients had applied for a job before presentation. Work resumption rates were similar across centers: 50 patients (61.7%) in Amsterdam, 42 (59.2%) in Munich, and 19 (65.5%) in Paris.
FIGURE 1.
Work status Sankey diagrams for patients of A, all centers; B, Amsterdam; C, Munich; and D, Paris.
TABLE 1.
Cohort Characteristics
| Characteristics | Amsterdam (N = 94) | Munich (N = 75) | Paris (N = 38) | Total (N = 207) | P value |
|---|---|---|---|---|---|
| Patient features | |||||
| Sex (male) | 52 (55.3%) | 37 (49.3%) | 25 (65.8%) | 114 (55.1%) | .251 |
| Age, mean (SD) | 39.5 (10.9) | 39.1 (10.2) | 43.7 (12.8) | 40.1 (11.1) | .152 |
| Presenting symptoms | |||||
| Seizure | 71 (75.5%) | 40 (53.3%) | 31 (81.6%) | 142 (68.6%) | .001b |
| Incidental | 10 (10.6%) | 10 (13.3%) | 5 (13.2%) | 25 (12.1%) | .845 |
| Headache | 7 (7.4%) | 12 (16.0%) | 0 (0.0%) | 19 (9.2%) | .015 |
| Sensory deficit | 6 (6.4%) | 9 (12.0%) | 0 (0.0%) | 15 (7.2%) | .061 |
| Visual disorders | 3 (3.2%) | 6 (8.0%) | 0 (0.0%) | 9 (4.3%) | .109 |
| Cognitive disorders | 1 (1.1%) | 5 (6.7%) | 0 (0.0%) | 6 (2.9%) | .049 |
| Motor deficit | 1 (1.1%) | 4 (5.3%) | 1 (2.6%) | 6 (2.9%) | .257 |
| Behavior change | 3 (3.2%) | 1 (1.3%) | 0 (0.0%) | 4 (1.9%) | .432 |
| Increased ICP | 0 (0.0%) | 2 (2.7%) | 1 (2.6%) | 3 (1.4%) | .282 |
| Vertigo | 0 (0.0%) | 3 (4.0%) | 0 (0.0%) | 3 (1.4%) | .069 |
| Aphasia | 1 (1.1%) | 1 (1.3%) | 0 (0.0%) | 2 (1.0%) | .784 |
| Fatigue | 1 (1.1%) | 0 (0.0%) | 0 (0.0%) | 1 (0.5%) | .547 |
| Inattention | 1 (1.1%) | 0 (0.0%) | 0 (0.0%) | 1 (0.5%) | .547 |
| Psychosis | 1 (1.1%) | 0 (0.0%) | 0 (0.0%) | 1 (0.5%) | .547 |
| Tumor features | |||||
| WHO gradea | .277 | ||||
| II | 50 (53.2%) | 43 (57.3%) | 26 (68.4%) | 119 (57.5%) | |
| III | 44 (46.8%) | 32 (42.7%) | 12 (31.6%) | 88 (42.5%) | |
| Volume, mL, mean (SD) | 63.2 (55.4) | 56.5 (112.7) | 49.9 (45.5) | 58.4 (78.4) | .036 |
| Hemisphere (L) | 51 (54.3%) | 29 (38.7%) | 22 (57.9%) | 102 (49.3%) | .118 |
| Treatment features | |||||
| Awake surgery | 75 (79.8%) | 11 (14.7%) | 31 (81.6%) | 117 (56.5%) | <.001b |
| EoR% (SD) | 80.3 (24.8) | 85.9 (19.5) | 76.1 (31.0) | 81.4 (24.6) | .049 |
| Chemotherapy | 29 (30.9%) | 53 (70.7%) | 9 (23.7%) | 91 (44.0%) | <.001b |
| Radiation therapy | 47 (50.0%) | 51 (68.0%) | 7 (18.4%) | 105 (50.7%) | <.001b |
| Work status | |||||
| Employed | 81 (86.2%) | 71 (94.7%) | 29 (76.3%) | 181 (87.4%) | .018 |
| Work resumption | 50 (61.7%) | 42 (59.2%) | 19 (65.5%) | 111 (61.3%) | .352 |
| ISCO major group | .293 | ||||
| 1 Managers | 8 (8.5%) | 6 (8.0%) | 8 (21.1%) | 22 (10.6%) | |
| 2 Professionals | 23 (24.5%) | 25 (33.3%) | 13 (34.2%) | 61 (29.5%) | |
| 3 Technicians | 10 (10.6%) | 11 (14.7%) | 2 (5.3%) | 23 (11.1%) | |
| 4 Clerical Support | 6 (6.4%) | 9 (12.0%) | 2 (5.3%) | 17 (8.2%) | |
| 5 Service and Sales | 12 (12.8%) | 5 (6.7%) | 5 (13.2%) | 22 (10.6%) | |
| 6 Skilled Agricultural, Forestry, and Fishery | 1 (1.1%) | 3 (4.0%) | 1 (2.6%) | 5 (2.4%) | |
| 7 Craft workers | 6 (6.4%) | 3 (4.0%) | 3 (7.9%) | 12 (5.8%) | |
| 8 Plant and Machine operators | 0 (0.0%) | 2 (2.7%) | 1 (2.6%) | 3 (1.4%) | |
| 9 Elementary | 6 (6.4%) | 2 (2.7%) | 0 (0.0%) | 8 (3.9%) | |
| ISCO skill level | .163 | ||||
| 1 | 6 (6.4%) | 2 (2.7%) | 0 (0.0%) | 8 (3.9%) | |
| 2 | 25 (26.6%) | 22 (29.3%) | 12 (31.6%) | 59 (28.5%) | |
| 3 | 13 (13.8%) | 12 (16.0%) | 2 (5.3%) | 27 (13.0%) | |
| 4 | 28 (29.8%) | 30 (40.0%) | 21 (55.3%) | 79 (38.2%) |
EoR, extent of resection; ICP, intracranial pressure; ISCO, International Standardized Classification of Occupations; WHO, World Health Organization.
As defined by the WHO 2016 classification.
P-value ≤.01.
Patient, Tumor, and Treatment Characteristics
The mean age was 39.5 years, 114 patients (55.1%) were male, and seizures were the most common symptom at presentation. Patient populations among hospitals were similar in age and sex. Seizures were less common in Munich (53.3%) compared with Amsterdam (75.5%) and Paris (81.6%), P = .001.
The distribution of tumor grades was not different between hospitals with 57.5% grade II tumors overall. Fewer patients in Munich underwent awake surgery (14.7%) compared with Amsterdam (79.8%) and Paris (81.6%), P < .001. The Munich cohort had more frequent adjuvant chemotherapy (70.7%) compared with Amsterdam (30.9%) and Paris (23.7%), P < .001. Less patients in Paris received adjuvant radiotherapy compared with the other hospitals (18.4% vs 50.0% in Amsterdamand 68.0% in Munich, P < .001).
Patient, Tumor, and Treatment Characteristics Potentially Associated With Work Resumption
In univariable logistic regression, male patients (odds ratio [OR] = 2.25, CI = 1.19-4.32, P = .013) and younger age (OR = 0.60, CI = 0.42-0.84, P = .004) were positively associated with work resumption. In addition, lower WHO grade (OR = 0.48, CI = 0.25-0.90), smaller tumor volume (OR = 0.40, CI = 0.21-0.73, P = .004), larger EoR (OR = 1.67, CI = 1.21-2.36, P = .003), and not receiving chemotherapy (OR = 0.45, CI = 0.23-0.85, P = .014) were related to resuming work after surgery (Table 2). These features were used as input in multivariable logistic regression.
TABLE 2.
Univariable Logistic Regression of Patient, Tumor, and Treatment Characteristics for Work Resumption
| Features | OR | 95% CI | P value |
|---|---|---|---|
| Patient features | |||
| Sex male (ref = female) | 2.25 | 1.19-4.32 | .013a |
| Age (per 10 years) | 0.60 | 0.42-0.84 | .004a |
| Hospital Munich (ref = Amsterdam) | 0.61 | 0.30-1.21 | .161 |
| Hospital Paris (ref = Amsterdam) | 0.89 | 0.35-2.35 | .803 |
| Presenting symptoms | |||
| Seizures | 0.54 | 0.25-1.11 | .101 |
| Tumor features | |||
| WHO grade III (ref = grade II) | 0.48 | 0.25-0.90 | .023a |
| Tumor volume (per 10 mL) | 0.40 | 0.21-0.73 | .004a |
| Tumor hemisphere right (ref = left) | 1.49 | 0.78-2.89 | .228 |
| Treatment features | |||
| Awake surgery | 1.27 | 0.67-2.40 | .462 |
| Extent of resection | 1.67 | 1.21-2.36 | .003a |
| Chemotherapy received | 0.45 | 0.23-0.85 | .014a |
| Radiotherapy received | 0.65 | 0.34-1.23 | .186 |
| Work status (ref = skill level 4) | |||
| ISCO skill level 1 | 0.42 | 0.08-2.03 | .276 |
| ISCO skill level 2 | 1.08 | 0.52-2.28 | .835 |
| ISCO skill level 3 | 1.43 | 0.54-4.10 | .485 |
ISCO, International Standardized Classification of Occupations; OR, odds ratio; WHO, World Health Organization.
P-value <.1.
Resected Regions Potentially Associated With Work Resumption
In imaging analysis, 4 patient scans failed registration quality checks with nonlinear registration but were approved after affine registration. Six patients were excluded because of insufficient imaging quality even after affine registration.
Figure 2 shows resection distributions for patient populations by work resumption status at follow-up. For comparison, preoperative tumor distribution maps are presented in Supplemental Digital Content 1 (http://links.lww.com/NS9/A39). The top 5 resected brain regions with potential importance for work resumption were the left limbic network, left acoustic radiation, left superior longitudinal fasciculus 3, the forceps major of the corpus callosum, and the left superior longitudinal fasciculus 2 according to the Bayesian hurdle model. The full difference posterior density distributions of regions are provided in Supplemental Digital Content 1 (http://links.lww.com/NS9/A39).
FIGURE 2.
Resection distribution maps of patients who had resumed work (top) vs not resumed work (bottom) at follow-up. The voxel intensity represents the amount of patients with a lesion in that specific voxel. The plotted numbers below refer to Montreal Neurological Institute z-axis slice. L, left; R, right.
Sex, Age, and Extent of Resection are Associated With Work Resumption
Male patients were more likely to resume work compared with female patients (OR = 2.87, CI = 1.36-6.26, P = .003) in multivariable logistic regression of selected characteristics. Younger patients were also more likely to resume work than older patients (OR = 0.56, CI = 0.36-0.83, P = .001). Finally, patients who had larger EoR (OR = 1.02, CI = 1.01-1.04, P = .004) were more likely to resume work (Table 3). WHO grade and adjuvant chemotherapy were no longer significantly associated with work resumption.
TABLE 3.
Multivariable Logistic Regression of Patient, Tumor, and Treatment Characteristics for Work Resumption
| Features | OR | 95% CI | P value |
|---|---|---|---|
| Patient features | |||
| Sex male (ref = female) | 2.87 | 1.36-6.26 | .007a |
| Age (per 10 years) | 0.56 | 0.36-0.83 | .005a |
| Tumor features | |||
| WHO grade III (ref = grade II) | 1.14 | 0.47-2.82 | .775 |
| Tumor volume (per 10 mL) | 0.44 | 0.20-0.91 | .036 |
| Treatment features | |||
| Extent of resection | 1.02 | 1.01-1.04 | .004a |
| Chemotherapy received | 0.46 | 0.19-1.08 | .077 |
OR, odds ratio; WHO, World Health Organization.
P-value ≤.01.
On adding lesion loads of selected regions, none were significantly associated with work resumption (Table 4) and model performance worsened: AIC = 182 without lesion load and AIC = 190 with lesion load.
TABLE 4.
Multivariable Logistic Regression of Patient, Tumor, and Treatment Characteristics and Cavity Lesion Load of Selected Regions for Work Resumption
| Features | OR | 95% CI | P value |
|---|---|---|---|
| Patient features | |||
| Sex male (ref = female) | 2.88 | 1.35-6.37 | .007a |
| Age | 0.56 | 0.36-0.85 | .008a |
| Tumor features | |||
| WHO grade III (ref = grade II) | 1.09 | 0.45-2.74 | .843 |
| Tumor volume (mL) | 0.51 | 0.21-1.04 | .114 |
| Treatment features | |||
| Extent of resection | 1.02 | 1.01-1.04 | .004a |
| Chemotherapy received | 0.41 | 0.17-1.00 | .051 |
| Cavity lesion load | |||
| Limbic network left | 0.81 | 0.42-1.54 | .519 |
| Acoustic radiation left | 0.96 | 0.50-1.85 | .901 |
| Superior longitudinal fasciculus 3 left | 1.13 | 0.43-3.13 | .803 |
| Forceps major | 1.07 | 0.74-1.74 | .719 |
| Superior longitudinal fasciculus 2 left | 0.81 | 0.31-2.09 | .661 |
OR, odds ratio; WHO, World Health Organization.
P-value ≤.01.
DISCUSSION
In this multicenter cohort study, we found no evidence that resection location was associated with work resumption. Of employed patients, 61% resumed work within the first year after surgery for lower-grade glioma. Younger age, male sex, and larger EoR were identified as independent predictors for work resumption.
At first neurosurgical consultation, 87% of patients were employed of which 61% resumed work within 1 year after surgery. This rate is in the lower range of previously reported rates of 52%, 71%, 81%, 85%, 82%, 90%, and 91%.3-5,7-10 The study reporting a lower work resumption rate may be dependent on socioeconomic factors such as elaborate sickness benefit schemes,4 whereas higher work resumption rates may be due to smaller cohort sizes and symptomatology bias in the selected patient cohort.3,7 Interestingly, we did find a substantial significant difference in the frequency of awake surgeries between centers which did not affect work resumption rate contrary to recent findings.10,15 These differences may be explained by selection bias given the cohort.
Our most salient finding is that we did not observe a relation between lesion load of resected regions and work resumption, suggesting that the location of surgery does not create a significant additional risk for work resumption. Model performance worsened on adding lesion loads indicating that work resumption is better explained by sex, age, and EoR rather than lesion load, which adds noise to the model. Surgical risk assessment and the application of perioperative techniques such as navigation, diffusion tensor imaging tracking, and awake mapping may effectively mitigate the risk of new neurological deficit, which could hamper work resumption. Although we did not find a relation with work resumption, awake mapping during surgery has been described to limit surgical risk.10,15,35 Another potential explanation for the lack of resected regions putting work resumption at risk could be that complex brain functions are supported by networks of regions distributed across the brain. Cognitive flexibility recovery depends on the extent of damage to the cognitive control network, and the threshold level of damage to this network can be observed in very different locations (frontal, parietal, temporal) on both sides of the brain.36,37 Finally, the definition of atlases and lesion measures may affect our findings. In future work, functional network measures from different scan modalities could provide a more holistic view.38-40
Lower EoR posed risk to work resumption, which has been described before.5,8,9 With lower EoR, tumor regrowth may occur sooner. In addition, larger EoR may only be achievable in less eloquent brain locations, thus posing less risk for postoperative deficits.
Younger age and male sex were also positive predictors for work resumption and have both been reported before.4,7 Older patients often have delayed recovery times and a higher complication risk in glioma surgery, negatively influencing functional outcome, which may hamper ability to return to work.41-44 Socioeconomic factors such as the motivation and necessity for resuming work might be lower for older patients because of financial independence or less financial responsibilities. In addition, men still participate more in the general labor force.45 A previous study found that being the sole household breadwinner is a positive predictor for work resumption, which is still more often true for men than women.8,46
In this study, we consider not returning to work as an adverse event. However, this view may not necessarily reflect reality. Imaginably, being diagnosed with an incurable cancer uproots patients' lives and influences how they value time spend at work.47 In addition, socioeconomic factors such as sickness benefit schemes may also influence the decision to return to work. In some instances, being determined unfit for work may benefit patients financially. Consulting patients on their wishes and discussing these values may help improve their quality of life.48
Strengths
This study is, to our knowledge, the first study to structurally analyze the relationship between location of surgical resection and work resumption after surgery for lower-grade glioma. Furthermore, it includes a heterogeneous data set from an international multicenter cohort making the results more generalizable. The inclusion of resection location combined with patient-, tumor-, and treatment-related characteristics is a major strength of our study and adds important information for patient consultation before surgery.
Limitations
Our study has some technical limitations. First, although the best suitable measure for our study, lesion load does not fully correspond to impact on functional regions. Second, lesion segmentations are subject to inter- and intrarater variability.49 Third, registration to standard space of lesioned brains is inherently difficult. We tried to limit influence of faulty registration by including manual checks at each image processing step.25 Fourth, resulting confidence intervals were broad. A larger cohort size may further specify results. Our study has data limitations as well. We had limited information on the number of working hours before and after surgery; decreased working hours may be related to patient's well-being.13 Furthermore, work resumption measures could be subject to timing. In multiple studies, work resumption rate increased in the second year after surgery.4,5,7 In addition, our data set lacked information on postoperative symptoms such as seizures and antiepileptic drug use. Having seizures may be a decisive factor of work resumption as driving (to work) and physical jobs may become impossible.5,14 Furthermore, extensive neuropsychological evaluation could provide more insight in the reasons for not returning to work. Finally, the data set may not represent more recent advances in awake mapping of cognitive functioning, which may affect work resumption.50
CONCLUSION
Resection location was not associated with work resumption after lower-grade glioma surgery. Male sex, younger age, and larger EoR were predictors for work resumption. As work resumption may rely heavily on socioeconomic and psychological factors, caution must be exercised when considering it as an outcome measure of surgery; nevertheless, it is a reasonable proxy of surgical outcome. Our results can be useful for patient counseling.
Acknowledgments
The authors thank Maximilian Schwendner, MD affiliated with neurosurgery departments of Heidelberg University Hospital and Klinikum rechts der Isar, Technische Universität München, for his help with data collection. Author Contributions: The conceptualization of this study was done by I.O.K., R.S.E., E.M., S.I., S.M.K., P.C.d.W.H. The methodology was designed by I.O.K., M.N.G.v.G., R.S.E., M.G.W., T.v.d.B, F.B., P.C.d.W.H. Data collection was executed by I.O.K., Y.d.H., E.M.,S.I., S.M.K. and data curation by I.O.K., M.N.G.v.G., R.S.E. Formal analysis was performed by I.O.K., M.N.G.v.G. The original draft of the manuscript was written by I.O.K., M.N.G.v.G, and manuscript review and editing was done by R.S.E., M.G.W., Y.d.H, T.v.d.B, E.M., S.I., S.M.K., F.B., P.C.d.W.H. All authors commented on previous drafts and approved the final manuscript. The study was supervised by R.S.E., F.B., P.C.d.W.H.
Footnotes
Presented at European Association of Neuro-Oncology meeting, Vienna, Austria, from September 15 to 18, 2022.
Supplemental digital content is available for this article at neurosurgerypractice-online.com.
Ivar O. Kommers and Maisa N. G. van Genderen contributed equally to this work.
Contributor Information
Ivar O. Kommers, Email: i.kommers@amsterdamumc.nl.
Maisa N. G. van Genderen, Email: m.n.g.vangenderen@amsterdamumc.nl.
Roelant S. Eijgelaar, Email: r.eijgelaar@amsterdamumc.nl.
Marnix G. Witte, Email: m.witte@nki.nl.
Yvette de Haan, Email: y.dehaan@erasmusmc.nl.
Tim van de Brug, Email: tvdbrug@gmail.com.
Emmanuel Mandonnet, Email: emmanuel.mandonnet@aphp.fr.
Sebastian Ille, Email: Sebastian.Ille@med.uni-heidelberg.de.
Sandro M. Krieg, Email: sandro.krieg@tum.de.
Frederik Barkhof, Email: f.barkhof@amsterdamumc.nl.
Funding
This work was funded by an unrestricted grant of Stichting Hanarth fonds, “Machine learning for better neurosurgical decisions in patients with glioblastoma”, and a grant for public-private partnerships (Amsterdam UMC PPP-grant) sponsored by the Dutch government through the Rijksdienst voor Ondernemend Nederland and Topsector Life Sciences and Health, “Picturing predictions for patients with brain tumors”. F.B. is funded by the National Institute for Health Research biomedical research centre at UCLH.
Disclosures
Sebastian Ille receives payments from Brainlab, Nexstim, and Carl Zeiss Meditec, and received travel cost payments in the past from Icotec AG. Sandro M. Krieg is a consultant for Brainlab AG. The other authors have no personal, financial, or institutional interest in any of the materials or devices described in this article.
SUPPLEMENTAL DIGITAL CONTENT
Supplemental Digital Content 1. Methods, 2 Figures. The Supplemental Digital Content expands on the Methods provided. Figure 1. Preoperative tumor distribution maps split according to work resumption at follow-up. Figure 2. Resection lesion load importance in all tested brain regions for work resumption.
REFERENCES
- 1.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 Oncol. 2023;25(12 Suppl 2):IV1-IV99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.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]
- 3.Birladeanu A, Barberis M, Poisson I, Froelich S, Mandonnet E. Oncological and functional outcomes support early resection of incidental IDH-mutated glioma. Acta Neurochir (Wien). 2023;165(10):2755-2767. [DOI] [PubMed] [Google Scholar]
- 4.Rydén I, Carstam L, Gulati S, et al. Return to work following diagnosis of low-grade glioma: a nationwide matched cohort study. Neurology. 2020;95(7):e856-e866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ng S, Herbet G, Moritz-Gasser S, Duffau H. Return to work following surgery for incidental diffuse low-grade glioma: a prospective series with 74 patients. Neurosurgery. 2020;87(4):720-729. [DOI] [PubMed] [Google Scholar]
- 6.Nicklin E, Phang I, Short SC, Hoogendoorn P, Boele FW. Patient and caregiver return to work after a primary brain tumor. Neurooncol Pract. 2023;10(6):565-575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Senft C, Behrens M, Lortz I, et al. The ability to return to work: a patient-centered outcome parameter following glioma surgery. J Neurooncol. 2020;149(3):403-411. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.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]
- 9.Muto J, Dezamis E, Rigaux-Viode O, et al. Functional-based resection does not worsen quality of life in patients with a diffuse low-grade glioma involving eloquent brain regions: a prospective cohort study. World Neurosurg. 2018;113:e200-e212. [DOI] [PubMed] [Google Scholar]
- 10.Barberis M, Poisson I, Prévost-Tarabon C, et al. Verbal fluency predicts work resumption after awake surgery in low-grade glioma patients. Acta Neurochir (Wien). 2024;166(1):88. [DOI] [PubMed] [Google Scholar]
- 11.Moritz-Gasser S, Herbet G, Maldonado IL, Duffau H. Lexical access speed is significantly correlated with the return to professional activities after awake surgery for low-grade gliomas. J Neurooncol. 2012;107(3):633-641. [DOI] [PubMed] [Google Scholar]
- 12.Robe PA, Rados M, Spliet WG, et al. Early surgery prolongs professional activity in IDH mutant low-grade glioma patients: a policy change analysis. Front Oncol. 2022;12:851803-851810. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Schiavolin S, Mariniello A, Broggi M, et al. Characteristics of patients returning to work after brain tumor surgery. Front Hum Neurosci. 2020;14:609080. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mandonnet E, De Witt Hamer P, Poisson I, et al. Initial experience using awake surgery for glioma: oncological, functional, and employment outcomes in a consecutive series of 25 cases. Neurosurgery. 2015;76(4):382-389. [DOI] [PubMed] [Google Scholar]
- 15.Duffau H. Awake mapping with transopercular approach in right insular-centered low-grade gliomas improves neurological outcomes and return to work. Neurosurgery. 2022;91(1):182-190. [DOI] [PubMed] [Google Scholar]
- 16.Beerda DCE, Zegers AD, van Andel ES, et al. Experiences and perspectives of patients with advanced cancer regarding work resumption and work retention: a qualitative interview study. Support Care Cancer. 2022;30(12):9713-9721. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Schmidt ME, Scherer S, Wiskemann J, Steindorf K. Return to work after breast cancer: the role of treatment-related side effects and potential impact on quality of life. Eur J Cancer Care (Engl). 2019;28(4):e13051. [DOI] [PubMed] [Google Scholar]
- 18.Bates E, Wilson SM, Saygin AP, et al. Voxel-based lesion-symptom mapping. Nat Neurosci. 2003;6(5):448-450. [DOI] [PubMed] [Google Scholar]
- 19.Rorden C, Fridriksson J, Karnath HO. An evaluation of traditional and novel tools for lesion behavior mapping. Neuroimage. 2009;44(4):1355-1362. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.von Elm E, Altman DG, Egger M, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344-349. [DOI] [PubMed] [Google Scholar]
- 21.ISCO-08. International Standard Classification of Occupations, Vol I. Department of Statistics International Labour Office; 2008. [Google Scholar]
- 22.De Witt Hamer PC, Hendriks EJ, Mandonnet E, Barkhof F, Zwinderman AH, Duffau H. Resection probability maps for quality assessment of glioma surgery without brain location bias. PLoS One. 2013;8(9):e73353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hendriks EJ, Habets EJJ, Taphoorn MJB, et al. Linking late cognitive outcome with glioma surgery location using resection cavity maps. Hum Brain Mapp. 2018;39(5):2064-2074. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fonov V, Evans AC, Botteron K, et al. Unbiased average age-appropriate atlases for pediatric studies. Neuroimage. 2011;54(1):313-327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Visser M, Petr J, Müller DMJ, et al. Accurate MR image registration to anatomical reference space for diffuse glioma. Front Neurosci. 2020;14:585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Avants BB, Tustison N, Johnson H. Advanced normalization tools (ANTS). Insight J. 2009;2(365):1-35. [Google Scholar]
- 27.Thomas Yeo BT, Krienen FM, Sepulcre J, et al. The organization of the human cerebral cortex estimated by intrinsic functional connectivity. J Neurophysiol. 2011;106(3):1125-1165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Makris N, Goldstein JM, Kennedy D, et al. Decreased volume of left and total anterior insular lobule in schizophrenia. Schizophr Res. 2006;83(2-3):155-171. [DOI] [PubMed] [Google Scholar]
- 29.Frazier JA Chiu S Breeze JL, et al. Structural brain magnetic resonance imaging of limbic and thalamic volumes in pediatric bipolar disorder. Am J Psychiatry. 2005;162(7):1256–1265. [DOI] [PubMed] [Google Scholar]
- 30.Desikan RS, Ségonne F, Fischl B, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006;31(3):968-980. [DOI] [PubMed] [Google Scholar]
- 31.Goldstein JM, Seidman LJ, Makris N, et al. Hypothalamic abnormalities in schizophrenia: sex effects and genetic vulnerability. Biol Psychiatry. 2007;61(8):935-945. [DOI] [PubMed] [Google Scholar]
- 32.Warrington S, Bryant KL, Khrapitchev AA, et al. XTRACT—standardised protocols for automated tractography in the human and macaque brain. Neuroimage. 2020;217:116923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Kruschke JK. Hierarchical Models. In: Doing Bayesian Data Analysis. Elsevier; 2015:221-264. [Google Scholar]
- 34.Carpenter B, Gelman A, Hoffman MD, et al. Stan: a probabilistic programming language. J Stat Softw. 2017;76(1):1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.De Witt Hamer PC, Robles SG, Zwinderman AH, Duffau H, Berger MS. Impact of intraoperative stimulation brain mapping on glioma surgery outcome: a meta-analysis. J Clin Oncol. 2012;30(20):2559-2565. [DOI] [PubMed] [Google Scholar]
- 36.Mrah S, Descoteaux M, Wager M, et al. Network-level prediction of set-shifting deterioration after lower-grade glioma resection. J Neurosurg. 2022;137(5):1329-1337. [DOI] [PubMed] [Google Scholar]
- 37.Weyer-Jamora C, Brie MS, Luks TL, et al. Cognitive impact of lower-grade gliomas and strategies for rehabilitation. Neurooncol Pract. 2021;8(2):117-128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sanai N, Berger MS. Glioma extent of resection and its impact on patient outcome. Neurosurgery. 2008;62(4):753-266. [DOI] [PubMed] [Google Scholar]
- 39.Grabowski MM, Recinos PF, Nowacki AS, et al. Residual tumor volume versus extent of resection: predictors of survival after surgery for glioblastoma. J Neurosurg. 2014;121(5):1115-1123. [DOI] [PubMed] [Google Scholar]
- 40.Wood JR, Green SB, Shapiro WR. The prognostic importance of tumor size in malignant gliomas: a computed tomographic scan study by the Brain Tumor Cooperative Group. J Clin Oncol. 1988;6(2):338-343. [DOI] [PubMed] [Google Scholar]
- 41.Corell A, Carstam L, Smits A, Henriksson R, Jakola AS. Age and surgical outcome of low-grade glioma in Sweden. Acta Neurol Scand. 2018;138(4):359-368. [DOI] [PubMed] [Google Scholar]
- 42.Gulati S, Jakola AS, Johannesen TB, Solheim O. Survival and treatment patterns of glioblastoma in the elderly: a population-based study. World Neurosurg. 2012;78(5):518-526. [DOI] [PubMed] [Google Scholar]
- 43.Ening G, Osterheld F, Capper D, Schmieder K, Brenke C. Risk factors for glioblastoma therapy associated complications. Clin Neurol Neurosurg. 2015;134:55-59. [DOI] [PubMed] [Google Scholar]
- 44.Kommers I, Ackermans L, Ardon H, et al. Between‐hospital variation in rates of complications and decline of patient performance after glioblastoma surgery in the Dutch Quality Registry Neuro Surgery. J Neurooncol. 2021;152(2):289-298. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Gender Statistics. Accessed 01, 2024. https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Gender_statistics
- 46.Coron C, Schmidt G. Sex, breadwinner status, and perceived job insecurity: a comparative analysis in Europe. Econ Ind Democracy. 2022;44(4):1052-1083. [Google Scholar]
- 47.Main DS, Nowels CT, Cavender TA, Etschmaier M, Steiner JF. A qualitative study of work and work return in cancer survivors. Psychooncology. 2005;14(11):992-1004. [DOI] [PubMed] [Google Scholar]
- 48.Zegers AD, Coenen P, Heeren A, et al. Work-related experiences and unmet needs of patients with a malignant glioma and relevant professionals: the BrainWork study. J Cancer Survivorship. 2023;19(1):326-338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Visser M, Müller DMJ, van Duijn RJM, et al. Inter-rater agreement in glioma segmentations on longitudinal MRI. Neuroimage Clin. 2019;22:101727. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Mandonnet E, Herbet G. Intraoperative Mapping of Cognitive Networks: Which Tasks for Which Locations. Springer International Publishing; 2021. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Digital Content 1. Methods, 2 Figures. The Supplemental Digital Content expands on the Methods provided. Figure 1. Preoperative tumor distribution maps split according to work resumption at follow-up. Figure 2. Resection lesion load importance in all tested brain regions for work resumption.