Low-grade gliomas (LGGs; WHO grade II) are primary brain tumors that on average grow slowly, and patients have a relatively favorable life expectancy. Yet, anaplastic transformation over time is inevitable.16 Surgery, and in particular the extent of resection (EOR), can significantly increase progression-free as well as overall survival of patients with LGG.10,13,14,35 Therefore, maximal safe tumor removal is currently considered the preferred first step in the neuro-oncological management of LGGs. Intraoperative stimulation mapping (ISM) has been shown to increase the EOR within so-called eloquent areas, while lowering the number of late severe neurological deficits.6,17 Most surgeries performed with awake conditions are nowadays still limited to monitoring language and/or motor functions, and typically do not involve assessment of the cognitive status of patients.26 Yet, preserving quality of life in patients with LGG implies more than maintaining classic neurological functions such as language and motricity. Cognitive functioning is important for quality of life, and affects the resumption of—among others—family, professional, and social activities after glioma surgery.1,7,12 Yet, the effect of surgery on cognitive performance remains debatable: new or worsened deficits (involving, among others, executive function, memory, and attention) have been reported in up to 55% of patients 6 months after surgery with ISM,3 but conversely, improvement of cognitive functioning is also reported.1,4 Although most surgical parameters (e.g., operation time, complications) seem comparable after surgery with or without ISM,6,7,17 potential differences in cognitive outcomes of patients with LGG undergoing surgery with versus without ISM have only been minimally examined.1,29
Decision-making during awake surgery in our practice is guided by the results of language and/or motor mapping, yet cognitive functioning is not assessed with formal tests during the procedure. However, we do routinely assess cognitive functioning before and after surgery in patients with LGG as part of clinical neuro-oncological care. These data were used to examine the effect of hemispheric tumor location and/or type of surgery (i.e., with or without ISM for motor and/or language mapping) on cognitive performance. Our first hypothesis was that cognitive deficits are present in patients with LGG irrespective of the hemispheric side of the lesion both before and after surgery. Since ISM procedures in our series typically did not involve extensive neuropsychological measures other than tests of language and/or motricity, our second hypothesis was that the use of ISM does not necessarily prevent declines in cognitive functions of patients with LGG.
Methods
Participants and Design
The present study was part of a larger prospective study in which patients with brain tumors admitted for craniotomy underwent preoperative (T0) and 3-month postoperative (T3) computerized neuropsychological assessment (NPA) implemented into clinical care at the Elisabeth-TweeSteden Hospital (Tilburg, The Netherlands). Cases eligible for the current study were patients who underwent initial resection of a histopathologically confirmed WHO grade II glioma (astrocytoma or oligodendroglioma) between November 2010 and March 2019. We did not include patients under the age of 18 years or those with a recent history (≤ 2 years) of severe psychiatric or neurological disorders, other major medical illnesses in the year prior to surgery, and/or a lack of basic proficiency in Dutch.
Dutch normative controls (NCs; age ranging from 20 to 80 years, 57% female, education ranging from 10 to 26 years) underwent NPA at baseline (n = 158) and at the 3-month follow-up (n = 131).23,24 Detailed information on the NC cohort is described in two former studies of our group.23,24 All participants provided written informed consent, and the study was approved by the Medical Ethics Committee.
Measures and Procedure
Sociodemographic Characteristics
NPAs were performed per standardized protocol, including a checklist and standardized interview during the first assessment (i.e., for obtaining sociodemographic information such as years of education).
Clinical Characteristics
Clinical information was obtained from the electronic medical charts. Tumor type was based on histopathological and molecular features and classified as either a WHO grade II astrocytoma or a WHO grade II oligodendroglioma. In addition, IDH-1 status (mutated/wild type) was classified, and (co)deletions of chromosome arms 1p and 19q were determined. By using preoperative gadolinium-enhanced T1-weighted MRI, we identified lesion side (i.e., right or left) and further classified tumor locations by frontal (i.e., frontal, frontal-temporal, frontal-parietal, frontal-insular) versus nonfrontal involvement. In addition, semiautomatic segmentations followed by manual adjustments on lesion margins were performed using ITK-SNAP (www.itksnap.org36) to determine preoperative tumor volume. EOR was classified by comparing T2-weighted and FLAIR signals on preoperative and 3-month postoperative scans as gross-total resection (complete resection of all T2 and FLAIR changes), near-total resection (91%–99% resection), subtotal resection (50%–90% resection), or partial resection (< 50% resection) independently by two investigators. Discrepancies were solved by consensus. The American Society of Anesthesiologists (ASA) score was considered as a measure of overall health.8 Patients with ASA scores I and II were considered healthy, whereas patients with scores III and IV were considered to have substantial comorbidities. Antiepileptic drugs and corticosteroids were recorded as psychotropic medication.
Functional MRI
We routinely used functional MRI (fMRI) with a verb generation task to assess hemispheric language dominance prior to surgery, using a cutoff value for a lateralization index (based on the number of active voxels in preselected regions of interest in both hemispheres) to classify a patient as having either typical language dominance (i.e., left-sided) or atypical language dominance (i.e., right-sided or bilateral).15,28
Surgical Procedures
In our practice, patients with LGG are surgically treated under awake conditions with ISM if the tumor is located in or near presumed critical regions for sensorimotor or language functions. The pre- and postcentral gyrus and associated corticospinal tract were considered critical regions for sensorimotor functions. Considered as critical regions for language functions were the left inferior frontal cortex, left (ventral) premotor cortex, and left posterior perisylvian cortex, as well as known subcortical language tracts—arcuate fasciculus, inferior fronto-occipital fasciculus, and fronto-aslant tract. We assume left-hemisphere language dominance in every patient, unless fMRI or clinical symptomatology suggests right-hemisphere dominance, as incidentally occurs, and we consequently also perform awake operations with ISM in these patients. In the current series, all ISM procedures in the right hemisphere were performed to monitor motor functions.
Patients in the ISM group underwent surgery under local anesthesia to allow for cortical and subcortical electrical stimulation of sensorimotor and language structures.27 A surgical guidance system (Stealth, Medtronic) was used with T1-weighted images with gadolinium, FLAIR images, and relevant subcortical fiber tracts from diffusion tensor imaging (corticospinal tract, inferior fronto-occipital fasciculus, arcuate fasciculus, and/or optic radiation). Tumor resection was performed according to the principles of sulcal-to-sulcal surgery whereby the EOR was tailored according to functional boundaries/patient performance as determined with ISM, and based on the surgical judgment that all FLAIR abnormalities had been removed (acknowledging limitations of surgical guidance systems due to brain shift). ISM was performed with a bipolar probe with 5-mm distance between the tips (ISIS, Inomed). A biphasic current was applied for a maximum duration of 4 seconds (frequency 60 Hz, pulse duration 0.5–1 msec, current 2–4 mA). During (sub)cortical stimulation as well as during tumor resection, patients performed language and/or motor tasks and were continuously monitored by a neuropsychologist.
Cognitive Performance
The preoperative and 3-month postoperative cognitive performance of patients was assessed using the official Dutch translation of the computerized battery Central Nervous System Vital Signs (CNS VS; https://cnsvs.com11). CNS VS comprises 7 neuropsychological tests (Table 1) that take approximately 30–40 minutes to complete. The local software application CNS VSX was used on a laptop. A well-trained test technician remained present.
Description of neuropsychological tests
| Test* | Content | Scores and Computation |
|---|---|---|
| Verbal memory test: verbal memory | Fifteen words presented one at a time. Subject subsequently identifies presented words among new words. | Total items correct |
| Visual memory test: visual memory | Fifteen abstract images presented one at a time. Subject subsequently identifies presented words among new words. | Total items correct |
| Finger tapping test: motor speed | Subject presses space bar as many times as possible in 10 secs with the index finger (3 trials per side). | Average taps right + average taps left |
| Symbol digit coding test: psychomotor speed | Participant matches numbers with symbols for 2 mins. | Correct responses − incorrect responses |
| Stroop test: interference | Part 1: subject presses space bar when a word is presented. Part 3: subject presses space bar if the color and meaning of the word do not match. | (Reaction time Part 3 − Reaction time Part 1)/Reaction time Part 1 |
| Shifting attention test: cognitive flexibility | Subject matches geometric objects by shape or color for 2 mins. | Correct responses − errors |
| Continuous performance test: sustained attention | Subject responds to target stimulus (B) among distractors for 5 mins. | Average reaction time of responses to B |
| Letter fluency test: verbal ability, executive control | Subject names words starting with a specific letter for 1 min (3 trials total). Three alternate test forms were used on subsequent assessments. | Total words correct |
The letter fluency test is a paper-and-pencil test; all others are CNS VS computerized tests.
Immediately after the battery is completed, a raw score for each test is generated by CNS VS. As effects of age, education, and sex were found on CNS VS performance in our group of Dutch NCs, raw scores were converted into sociodemographically adjusted z scores for each neuropsychological test.23 Moreover, the scores of patients at T3 are corrected for the effects of practice in addition to the sociodemographic corrections, because NCs showed significant practice effects between a baseline and follow-up assessment.24
Verbal processes were assessed with a Dutch version of the Controlled Oral Word Association (COWA) verbal letter fluency test, through naming as many words as possible beginning with a specific letter of the alphabet in 60 seconds, for a total of 3 letters in 3 minutes (D-A-T at T0, K-O-M at T3). The total number of correct words was converted into an education-corrected z score (no age or sex effects were demonstrated in Dutch NCs).32
Psychological Characteristics
As a part of the NPA, anxiety and depression were assessed by the Dutch translation of the Hospital Anxiety and Depression Scale (HADS).33,37 The HADS is a 14-item self-reported questionnaire with 2 scales measuring symptoms of anxiety (7 items) and depression (7 items). Response options range from 0 to 3, resulting in a score range from 0 to 21 for each subscale.
Statistical Analyses
Patient Characteristics
Descriptive and comparative analyses were performed using 2-sided independent t-tests and chi-square tests of independence. Baseline sociodemographic (age, sex, and educational level); clinical (tumor type, IDH-1 and 1p/19q status, tumor location, lesion side, tumor volume, epilepsy, ASA score, and medication use); and psychological (anxiety and depression) variables were analyzed in patients with right- and left-sided lesions, and in patients with versus without ISM.
Group and Time Effects on Cognitive Performance
Effects of group and time (i.e., T0–T3) were examined using a series (for each neuropsychological test score) of 2-way mixed ANOVAs. Groups for the first series of ANOVAs comprised patients with right-sided lesions, patients with left-sided lesions, and NCs; for the second series the groups comprised patients with ISM, patients without ISM, and NCs.
Interaction Effects
Interaction effects resulting from the ANOVAs indicated whether there was an interaction between the group and time. In the case of statistically significant interaction terms, simple main effects for group (1-way ANOVAs) and time (paired-samples t-tests) were performed to assess which group contributed to the significant effect.
Main Group Effects
Main group effects indicated whether the performance of patients differed from the performance of NCs, and whether differences in cognitive performance existed between patient groups (irrespective of time). Pairwise comparisons were interpreted in the case of significant main group effects to determine which group contributed to the overall significant main group effect.
Main Time Effects
Cognitive performance over time (T0–T3, irrespective of group) was also assessed using the ANOVAs. In case of statistically significant main time effects, pairwise comparisons were interpreted to determine which group contributed to the overall significant main time effect.
Partial eta-square was considered as a measure of effect size (ES), with values of 0.02 representing small effects; 0.13, medium effects; and 0.26, a large ES.5
Individual Changes in Performance Over Time
Reliable change indices (RCIs) were calculated for individual patients for each neuropsychological test score in order to determine whether observed changes in scores reliably reflect true performance changes, while taking into account methodological confounders such as practice effects. RCI formulae used are based on repeated assessments in our group of NCs.24,32 Change was defined as RCI values exceeding ± 1.645, which equates to the 90% CI, with positive values representing improvement and negative values representing a decline in performance. The number and percentages of patients with improved, stable, or a decline in performance were counted for each test for patients with right- versus left-sided lesions and for patients with ISM versus those without ISM, and then compared using chi-square tests of independence.
Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS) version 24.0 (IBM Corp.). To reduce the false discovery rate due to multiple statistical testing, p values were set against a corrected alpha, using the Benjamini-Hochberg (BH) procedure.2
Results
Patient Characteristics
Figure 1 shows the flow of patients with LGG in the current study. At baseline 92 patients were neuropsychologically assessed, and 77 patients completed both T0 and T3. Table 2 presents sociodemographic, clinical, and psychological characteristics of the study groups.
Flowchart of patients with LGG undergoing surgery and NPA. Figure is available in color online only.
Characteristics of patients with LGG
| Characteristic | Right-Sided Lesion, n = 38 | Left-Sided Lesion, n = 39 | With ISM, n = 42 | Without ISM, n = 35 |
|---|---|---|---|---|
| Age in yrs, mean ± SD (range) | 41.7 ± 13.7 (19–68) | 40.9 ± 11.9 (21–66) | 41.4 ± 11.9 (19–64) | 41.1 ± 13.9 (20–68) |
| Education in yrs, mean ± SD (range) | 14.8 ± 3.0 (9–21) | 15.8 ± 2.8 (11–21) | 15.6 ± 2.9 (10–21) | 15.03 ± 3.0 (9–21) |
| Sex: F/M | 18 (47)/20 (53) | 16 (41)/23 (59) | 18 (43)/24 (57) | 16 (46)/19 (54) |
| Astrocytoma II/oligodendroglioma II | 24 (63)/14 (37) | 24 (62)/15 (38) | 27 (64)/15 (36) | 21 (60)/14 (40) |
| Mutated/wild-type IDH-1* | 32 (86)/5 (14) | 34 (89)/4 (11) | 38 (93)/3 (7) | 28 (82)/6 (18) |
| 1p/19q: non–co-deleted/co-deleted† | 11 (42)/15 (58) | 17 (53)/15 (47) | 19 (56)/15 (44) | 9 (38)/15 (62) |
| Surgical procedure: with/without ISM | 11 (29)/27 (71) | 31 (80)/8 (20) | — | — |
| Frontal/nonfrontal tumor location | 27 (71)/11 (29) | 30 (77)/9 (23) | 31 (74)/11 (26) | 26 (74)/9 (26) |
| Right-sided lesion/left-sided lesion | — | — | 11 (26)/31 (74) | 27 (77)/8 (23) |
| Tumor vol in cm3, median (range)‡ | 41.6 (3.2–203.3) | 46.2 (1.3–226.9) | 50.8 (2.5–226.9) | 33.9 (1.3–203.3) |
| EOR: GTR/NTR/STR/PR§ | 4 (12)/14 (41)/13 (38)/3 (9) | 2 (6)/11 (31)/20 (57)/2 (6) | 4 (11)/11 (30)/19 (51)/3 (8) | 2 (6)/14 (44)/14 (44)/2 (6) |
| fMRI-determined language lateralization: typical/atypical¶ | 25 (89)/3 (11) | 33 (92)/3 (8) | 38 (91)/4 (9) | 20 (91)/2 (9) |
| Adjuvant therapy: none/RT/RT+chemo | 28 (74)/8 (21)/2 (5) | 30 (77)/8 (21)/1 (3) | 31 (79)/8 (19)/1 (2) | 25 (71)/8 (23)/2 (6) |
| Presentation with epilepsy: yes/no | 24 (63)/14 (37) | 27 (69)/12 (31) | 29 (69)/13 (31) | 22 (63)/13 (37) |
| ASA score: I + II/III + IV | 38 (100)/0 (0) | 37 (95)/2 (5) | 41 (98)/1 (2) | 34 (97)/1 (3) |
| Psychotropic medication: yes/no | 28 (74)/10 (26) | 36 (92)/3 (8) | 36 (86)/6 (14) | 28 (80)/7 (20) |
| Anxiety HADS, mean ± SD (range)** | 7.7 ± 3.8 (2–14) | 6.6 ± 3.8 (0–18) | 6.5 ± 3.7 (1–18) | 6.6 ± 4.2 (0–14) |
| Depression HADS, mean ± SD (range)** | 5.2 ± 3.8 (0–12) | 4.1 ± 3.2 (0–14) | 4.7 ± 3.6 (0–14) | 4.6 ± 3.4 (0–12) |
Chemo = chemotherapy; GTR = gross-total resection; NTR = near-total resection; PR = partial resection; RT = radiotherapy; STR = subtotal resection.
Unless otherwise indicated, values are expressed as the number of patients (%).
Data missing in 2 cases.
Data missing in 19 cases (determination since recent years).
Data missing in 9 cases due to missing/unusable scans.
Data missing in 8 cases due to missing/unusable scans.
Data missing in 13 cases due to lack of verb generation or fMRI data.
Data missing in 5 cases.
There were no significant differences regarding sociodemographic, clinical, and psychological characteristics between patients with right- or left-sided lesions (p > 0.073, for all comparisons), except for surgical procedure (p < 0.001). More patients with left-sided lesions underwent surgery with ISM, whereas more patients with right-sided lesions were operated on without ISM. Along these lines, there were no significant differences between patients with or without ISM (p > 0.155, all comparisons), except for the lesion side (p < 0.001). The EOR did not significantly differ after surgery with or without ISM, nor between patients with right- or left-sided lesions (p > 0.441, all comparisons).
Group and Time Differences in Cognitive Performance
Hemispheric Lesion Side
Interaction Effects.
We found a significant interaction between group (patients with right-sided lesions, patients with left-sided lesions, and NCs) and time (T0–T3) for the continuous performance test [F(2,200) = 3.69, p = 0.027, partial η2 = 0.04] (Table 3, Fig. 2). Post hoc tests showed a significantly lower T0 performance of patients with right-sided (but not with left-sided) lesions as compared to NCs, and lower T3 performance in both patient groups as compared to NCs. Furthermore, post hoc tests revealed significantly lower T3 performance (as compared to T0) in patients with left-sided lesions.
Cognitive performance of patients with LGG over time: group, time, and interaction effects
| Right-Sided/Left-Sided/NCs | With ISM/Without ISM/NCs | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Group Effects | Effects Over Time (T0–T3) | Interaction (group × time) | Group Effects | Effects Over Time (T0–T3) | Interaction (group × time) | |||||||||||||
| Cognitive Test* | F | p Value | Partial η2† | F | p Value | Partial η2† | F | p Value | Partial η2† | F | p Value | Partial η2† | F | p Value | Partial η2† | F | p Value | Partial η2† |
| Verbal memory test | 7.75 | 0.001‡ | 0.08 | 7.93 | 0.005‡ | 0.04 | 0.80 | 0.452 | 0.01 | 7.64 | <0.001‡ | 0.07 | 7.54 | 0.007‡ | 0.04 | 0.78 | 0.461 | 0.01 |
| Visual memory test | 3.52 | 0.031 | 0.04 | 4.42 | 0.037 | 0.02 | 0.68 | 0.508 | 0.01 | 2.51 | 0.084 | 0.03 | 4.99 | 0.027‡ | 0.03 | 0.62 | 0.541 | 0.01 |
| Finger tapping test | 10.67 | <0.001‡ | 0.10 | 1.36 | 0.245 | 0.01 | 2.02 | 0.135 | 0.02 | 9.76 | <0.001‡ | 0.09 | 1.39 | 0.240 | 0.01 | 0.97 | 0.380 | 0.01 |
| Symbol digit coding test | 9.72 | <0.001‡ | 0.09 | 3.75 | 0.054 | 0.02 | 0.06 | 0.941 | 0.00 | 11.35 | <0.001‡ | 0.10 | 3.37 | 0.068 | 0.02 | 0.40 | 0.672 | 0.00 |
| Stroop test | 2.15 | 0.119 | 0.02 | 0.31 | 0.577 | 0.02 | 2.32 | 0.101 | 0.02 | 4.99 | 0.008‡ | 0.05 | 0.13 | 0.720 | 0.00 | 0.94 | 0.394 | 0.01 |
| Shifting attention test | 5.93 | 0.003‡ | 0.06 | 7.51 | 0.007‡ | 0.04 | 1.19 | 0.307 | 0.01 | 8.02 | <0.001‡ | 0.07 | 7.22 | 0.008‡ | 0.04 | 2.07 | 0.129 | 0.02 |
| Continuous performance test | 15.06 | <0.001‡ | 0.13 | 11.36 | 0.001‡ | 0.05 | 3.69 | 0.027‡ | 0.04 | 14.77 | <0.001‡ | 0.13 | 9.88 | 0.002‡ | 0.05 | 5.91 | 0.003‡ | 0.06 |
| Verbal fluency test | 8.65 | <0.001‡ | 0.09 | 1.94 | 0.165 | 0.01 | 2.57 | 0.079 | 0.03 | 9.68 | <0.001‡ | 0.10 | 1.47 | 0.227 | 0.01 | 6.16 | 0.003‡ | 0.06 |
The verbal fluency test is a paper-and-pencil test; all others are CNS VS computerized tests.
Partial eta-square ES values of 0.02 represent small effects; 0.13, medium effects; and 0.26, large effects (Cohen).
p < BH-corrected alpha 0.03.
Graphs showing mean levels of performance (in terms of z scores, with 95% CIs) on neuropsychological tests of patients with LGG with right- versus left-sided lesions at pre- (T0) and postsurgery (T3). Figure is available in color online only.
Main Group Effects.
Main group analyses showed significant differences in mean performance between groups on 5 neuropsychological tests (Table 3). Pairwise comparisons showed that performance was lower in all patients (irrespective of lesion side) as compared to NCs on tests of shifting attention and verbal fluency (ES 0.06 and 0.09). Also, performance was lower in patients with right-sided (but not with left-sided) lesions as compared to NCs on verbal memory, finger tapping, and symbol digit coding (ES ranging from 0.08 to 0.10). There were no significant differences between patients with left- and right-sided lesions with regard to neuropsychological test scores.
Main Time Effects.
Significantly different mean levels of performance over T0–T3 were found for the verbal memory and shifting attention test. Post hoc tests revealed significantly lower performance on verbal memory in patients with left-sided lesions at T3 as compared to T0, but no significant differences over time on the shifting attention test for any of the groups.
Intraoperative Stimulation Mapping
Interaction Effects.
We found statistically significant interactions between group (ISM patients, non-ISM patients, and NCs) and time with regard to performance on the continuous performance test [F(2,202) = 5.91, p < 0.003, partial η2 = 0.06] and verbal fluency test [F(2,182) = 6.16, p = 0.003, partial η2 = 0.06] (Table 3, Fig. 3). Post hoc tests showed significantly lower performance on the continuous performance (T0 and T3) and verbal fluency (T0) test in both patient groups as compared to the NCs (p < 0.038, all comparisons). Also, postoperative performance on the verbal fluency test was lower in patients with ISM as compared to patients without ISM and NCs (p < 0.001). Post hoc tests revealed significantly lower performance at T3 as compared to T0 on the continuous performance test in patients with ISM, yet no significant effects of time were found for one of the groups on verbal fluency performance (p > 0.065, all comparisons).
Graphs showing mean levels of performance (in terms of z scores, with 95% CIs) on neuropsychological tests of patients with LGG undergoing surgery with or without ISM at pre- (T0) and postsurgery (T3). Figure is available in color online only.
Main Group Effects.
With regard to the remaining neuropsychological tests (i.e., for which no significant interaction effects were found), significant main group effects (irrespective of time point) were found for 5 of 6 tests (i.e., not for visual memory). Pairwise comparison tests revealed worse mean performance of both patient groups as compared to NCs on the finger tapping test, and worse performance of patients with ISM (but not of patients without ISM) as compared to NCs on the verbal memory, symbol digit coding test, and shifting attention test. Furthermore, significant differences between the two patient groups were demonstrated: patients with ISM performed significantly worse on the Stroop and shifting attention tests than did patients without ISM.
Main Time Effects.
Significant effects of time on performance on verbal and visual memory tests, and also on the shifting attention test, were demonstrated. Pairwise comparisons revealed significantly lower performance at T3 (as compared to T0) on the shifting attention test in patients with ISM (p = 0.018), but no significant differences with regard to verbal and visual memory were found over time for one of the groups, specifically.
Individual Changes in Performance Over Time
The percentages of patients per RCI category (i.e., declined, stable, or improved) are shown in Fig. 4. We found a decline in performance in 0%–19% of the patients with right-sided lesions and in 3%–25% of the patients with left-sided lesions over different neuropsychological tests. Improvement of performance over time was demonstrated by 3%–12% and 0%–11% of patients with right- and left-sided lesions, respectively. Five percent to 27% of the patients with ISM and 3%–15% of the patients without ISM showed a decline in performance. Zero percent to 12% of the patients with ISM showed improvement, and 0%–15% of the patients without ISM showed improvement.
Bar graph showing changes in cognitive performance of individuals with LGG. Figure is available in color online only.
We found no significant differences between the proportion of patients with improved, stable, or decreased performance over time in patients with right- versus left-sided lesions (p > 0.242, all comparisons), nor in patients with versus those without ISM (p > 0.107, all comparisons) on any of the neuropsychological tests.
Discussion
Our goal was to evaluate cognitive functioning of patients with LGG from pre- to postsurgery, and specifically to investigate the effect of hemispheric tumor location and surgical procedure (i.e., with or without motor/language mapping) on cognitive functioning. It is important to note that cognitive functions other than language were not formally tested during ISM in our series. Consistent with the literature, we found mildly to moderately impaired performance of patients with LGG both prior to and after surgery on the majority of the neuropsychological tests that target cognitive functions.23,30
Patients with right- and left-sided lesions scored significantly lower on 6 and 3 of 8 tests, respectively, as compared to NCs. No significant differences between performances of patients with right- versus left-sided lesions were found at baseline or at 3-month follow-up. We found, however, a significant deterioration of the mean performance of patients with left-sided lesions on tests of verbal memory and sustained attention. With regard to the surgical procedure, patients undergoing surgery with ISM versus those without ISM scored significantly lower on 6 and 3 of 8 tests, respectively, as compared to NCs. Patients with ISM performed worse than non-ISM patients on 3 tests (i.e., Stroop, shifting attention, and verbal fluency tests), and also ISM patients showed a significant deterioration of performance on the shifting attention and continuous performance tests over time.
Regarding changes in individual patients, RCIs demonstrated improved performance over time in at most 15% of the patients over the different neuropsychological tests. However, a decline in performance was found in up to 27% of the patients over different tests. Overall, patients showed the greatest deficits (on the mean group level) as well as the most deteriorations (on the individual patient level) of performance on tasks of memory, cognitive flexibility, attention, and verbal ability and executive functioning. This is in line with earlier studies on cognitive functioning of patients with LGG, mostly demonstrating, but not restricted to, language and higher-order executive functioning deficits.22,30
These results suggest that patients with LGG suffer from widespread cognitive impairments, irrespective of the hemispheric location of the tumor. This is in line with more recent clinical concepts that functions are represented in large-scale networks that span both hemispheres,25 and with a number of studies that found cognitive and behavioral deficits after glioma surgery in patients with right-sided lesions.20,25,34 Second, we found that ISM patients did not perform better than non-ISM patients (either before or after surgery), and even performed worse on the Stroop and shifting attention tests (irrespective of time). As we observed a correlation between scores on the tests of shifting attention and verbal fluency performance (data not shown), this posits that attentional performance of ISM patients might interfere with language performance. Vice versa, language performance may also bias performance on tests of cognitive flexibility. With regard to the shifting attention test (used in the current study to assess cognitive flexibility), patients are instructed to match geometric objects by either shape or color, whereby the rules (match shape/color) change at random. Although this test is not intended to assess verbal abilities, following the changing rules on the computer screen also relies on verbal processing abilities—the same holds true for the Stroop test. Worse performance of ISM patients as compared to non-ISM patients on these tests might therefore also reflect (in part) language difficulties. Along the same lines, worse cognitive performance of ISM patients might be a consequence of the selection criteria used to decide on surgery with or without ISM. Patients were considered for surgery with ISM if the LGG was located close to or infiltrated among other eloquent (language) areas, and are therefore more likely to show language deficits, which might bias or even affect performance on neuropsychological tests concerning functions other than language. Third, lower postoperative (as compared to preoperative) performance of patients with left-sided lesions and patients with ISM was found on verbal memory and sustained attention, and on cognitive flexibility and sustained attention, respectively. The non-ISM patients did not show significant deteriorations of mean performance over time.
Perhaps, awake surgery with only language and/or motor mapping offers a false sense of security for cognitive functions. When a patient remains cooperative and neurologically intact during surgery, we are generally encouraged to maximize the resection. However, he or she might (unknowingly) already have failed on other more specific neuropsychological tests, indicating that mapping of only language and/or motor functions is insufficient to prevent postoperative cognitive deficits. Future studies should investigate whether it is possible to preserve cognitive functioning following surgery with the use of ISM and specific tests addressing higher cognitive functions. Two recent studies suggested favorable effects of ISM with intraoperative neuropsychological testing with regard to the preservation of networks underlying executive performance;19,21 however, small numbers of patients (with heterogeneous glioma histologies) were included, and a control group without ISM was lacking.
We should be careful, however, because extensive intraoperative neuropsychological testing might lead to a prematurely terminated resection with oncological disadvantages. Future studies should consider broader neuropsychological testing during surgery with ISM in a systematic manner driven by a hypothesis based on current neuroscientific literature, whereby we strongly advocate to study not only the immediate (intraoperative) effects of cognitive performance during surgery but also the long-term effects on cognitive and socioprofessional functioning. Ideally, the resection should be stopped once these latter abilities are endangered. We do acknowledge that it will be challenging to find a cognitive test that can be performed by a patient during surgery and that at the same time can adequately reflect cognitive abilities in daily life.
Unexpected consequences of a surgical procedure can also provide useful insight into the functional neuroanatomy of cognition. Mandonnet intraoperatively tested for semantic memory and spatial neglect, yet observed significant and lasting deteriorations of set-shifting abilities (e.g., cognitive flexibility) after awake LGG surgery in the right hemisphere (presumably due to a surgical disconnection of the superior longitudinal fasciculus and arcuate fasciculus).18 Consequently, Mandonnet suggested that more specific tests should be added to address other cognitive functions. It is currently still unclear whether or not intraoperative tests can be developed that can specifically target critical structures for cognitive abilities (i.e., of functions that do not tend to recover after surgery when damaged), because even our current ISM procedures for language and/or motor functions do not necessarily protect against permanent functional loss.31 Deteriorations are usually mild and transient in clinical terms, and are therefore unlikely to be explained by damage to the larger language pathways (such as the arcuate fasciculus or inferior fronto-occipital fasciculus).
The current study has some limitations that should be noted. Patients undergoing surgery with versus without ISM were not individually matched or randomly allocated. We did, however, assess differences in sociodemographic, clinical, and psychological characteristics of patients between the two groups. No significant differences were found, with the exception of the lesion side, which is an inherent feature of the current neurosurgical management of patients with LGG. A subsample of patients was already receiving adjuvant radiotherapy and/or chemotherapy at the time of the 3-month follow-up assessment, and these therapies are known to increase the risk for cognitive impairments.4,31 Yet, as the vast majority of the patients did not yet receive adjuvant treatment and the portion of patients with radiotherapy and/or chemotherapy did not differ between hemispheric or surgical groups, it is unlikely that this factor led to biased results. Furthermore, we were not able to include lesion side/surgical procedure as covariants in the statistical analyses, because this would nearly result in a correction for the difference in conditions.9 Second, with regard to the mixed ANOVAs, we found several nonsignificant post hoc tests of time effects, whereas the global model was significant (i.e., suggesting an overall decline of performance that is not attributable to one of the patient groups). Last, although a statistical correction (i.e., the BH procedure) was applied to warrant the interpretation of false-positive results due to multiple statistical comparisons, replication research in other samples of patients with LGG with regard to the results of the current explorative study are necessary.
Conclusions
Patients with LGG showed widespread cognitive impairments both before and after surgery, irrespective of the hemispheric tumor location. Significant deteriorations of performance in patients with left-sided lesions and patients operated on with ISM were present in a subset of cognitive functions. We conclude that there is room for improvement of cognitive functioning in surgically treated patients with LGG. ISM of cognitive functions by using specific cognitive tests during surgery may potentially improve functional outcome, but that is to be determined in future studies and balanced against oncological outcome. Furthermore, the implementation of routine NPA in the clinical management of patients with LGG should be encouraged, to inform and alert patients and clinicians on the status of cognitive functioning.
Acknowledgments
This study was funded by ZonMw, a Dutch organization for Health Research and Development (project no. 842003007).
Disclosures
The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.
Author Contributions
Conception and design: Gehring, Sitskoorn, Rutten. Acquisition of data: Rijnen. Analysis and interpretation of data: Rijnen, Kaya, Gehring, Sitskoorn, Rutten. Drafting the article: Rijnen, Rutten. Critically revising the article: Rijnen, Kaya, Gehring, Sitskoorn, Rutten. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Rijnen. Statistical analysis: Rijnen. Administrative/technical/material support: Rijnen, Kaya. Study supervision: Gehring, Rutten.
References
- 1↑
Barzilai O, Ben Moshe S, Sitt R, Sela G, Shofty B, Ram Z: Improvement in cognitive function after surgery for low-grade glioma. J Neurosurg 130:426–434, 2019
- 2↑
Benjamini Y, Hochberg Y: Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B 57:289–300, 1995
- 3↑
Brennum J, Engelmann CM, Thomsen JA, Skjøth-Rasmussen J: Glioma surgery with intraoperative mapping—balancing the onco-functional choice. Acta Neurochir (Wien) 160:1043–1050, 2018
- 4↑
Campanella F, Palese A, Del Missier F, Moreale R, Ius T, Shallice T, et al.: Long-term cognitive functioning and psychological well-being in surgically treated patients with low-grade glioma. World Neurosurg 103:799–808.e9, 2017
- 5↑
Cohen J: Statistical Power Analysis for the Behavioral Sciences, ed 2. Hillsdale, NJ: Lawrence Erlbaum Associates, 1988
- 6↑
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 30:2559–2565, 2012
- 7↑
Drewes C, Sagberg LM, Jakola AS, Solheim O: Perioperative and postoperative quality of life in patients with glioma—a longitudinal cohort study. World Neurosurg 117:e465–e474, 2018
- 10↑
Garcia CR, Slone SA, Pittman T, St Clair WH, Lightner DD, Villano JL: Comprehensive evaluation of treatment and outcomes of low-grade diffuse gliomas. PLoS One 13:e0203639, 2018
- 11↑
Gualtieri CT, Johnson LG: Reliability and validity of a computerized neurocognitive test battery, CNS Vital Signs. Arch Clin Neuropsychol 21:623–643, 2006
- 12↑
Herbet G, Moritz-Gasser S: Beyond language: mapping cognition and emotion. Neurosurg Clin N Am 30:75–83, 2019
- 13↑
Hervey-Jumper SL, Berger MS: Evidence for improving outcome through extent of resection. Neurosurg Clin N Am 30:85–93, 2019
- 14↑
Jakola AS, Skjulsvik AJ, Myrmel KS, Sjåvik K, Unsgård G, Torp SH, et al.: Surgical resection versus watchful waiting in low-grade gliomas. Ann Oncol 28:1942–1948, 2017
- 15↑
Janecek JK, Swanson SJ, Sabsevitz DS, Hammeke TA, Raghavan M, Mueller W, et al.: Naming outcome prediction in patients with discordant Wada and fMRI language lateralization. Epilepsy Behav 27:399–403, 2013
- 16↑
Louis DN, Perry A, Reifenberger G, von Deimling A, Figarella-Branger D, Cavenee WK, et al.: The 2016 World Health Organization classification of tumors of the central nervous system: a summary. Acta Neuropathol 131:803–820, 2016
- 17↑
Lu VM, Phan K, Rovin RA: Comparison of operative outcomes of eloquent glioma resection performed under awake versus general anesthesia: a systematic review and meta-analysis. Clin Neurol Neurosurg 169:121–127, 2018
- 18↑
Mandonnet E: A surgical approach to the anatomo-functional structure of language. Neurochirurgie 63:122–128, 2017
- 19↑
Motomura K, Chalise L, Ohka F, Aoki K, Tanahashi K, Hirano M, et al.: Neurocognitive and functional outcomes in patients with diffuse frontal lower-grade gliomas undergoing intraoperative awake brain mapping. J Neurosurg [epub ahead of print May 17, 2019. DOI: 10.3171/2019.3.JNS19211]
- 20↑
Noll KR, Weinberg JS, Ziu M, Benveniste RJ, Suki D, Wefel JS: Neurocognitive changes associated with surgical resection of left and right temporal lobe glioma. Neurosurgery 77:777–785, 2015
- 21↑
Puglisi G, Sciortino T, Rossi M, Leonetti A, Fornia L, Conti Nibali M, et al.: Preserving executive functions in nondominant frontal lobe glioma surgery: an intraoperative tool. J Neurosurg 131:474–480, 2019
- 22↑
Racine CA, Li J, Molinaro AM, Butowski N, Berger MS: Neurocognitive function in newly diagnosed low-grade glioma patients undergoing surgical resection with awake mapping techniques. Neurosurgery 77:371–379, 2015
- 23↑
Rijnen SJM, Meskal I, Emons WHM, Campman CAM, van der Linden SD, Gehring K, et al.: Evaluation of normative data of widely used computerized neuropsychological battery: applicability and effects of sociodemographic variables in a Dutch sample. Assessment [epub ahead of print], 2017
- 24↑
Rijnen SJM, van der Linden SD, Emons WHM, Sitskoorn MM, Gehring K: Test-retest reliability and practice effects of a computerized neuropsychological battery: a solution-oriented approach. Psychol Assess 30:1652–1662, 2018
- 25↑
Rolland A, Herbet G, Duffau H: Awake surgery for gliomas within the right inferior parietal lobule: new insights into the functional connectivity gained from stimulation mapping and surgical implications. World Neurosurg 112:e393–e406, 2018
- 26↑
Ruis C: Monitoring cognition during awake brain surgery in adults: a systematic review. J Clin Exp Neuropsychol 40:1081–1104, 2018
- 27↑
Rutten GJ: Speech hastening during electrical stimulation of left premotor cortex. Brain Lang 141:77–79, 2015
- 28↑
Rutten GJ, Ramsey NF, van Rijen PC, Alpherts WC, van Veelen CW: FMRI-determined language lateralization in patients with unilateral or mixed language dominance according to the Wada test. Neuroimage 17:447–460, 2002
- 29↑
Sacko O, Lauwers-Cances V, Brauge D, Sesay M, Brenner A, Roux FE: Awake craniotomy vs surgery under general anesthesia for resection of supratentorial lesions. Neurosurgery 68:1192–1199, 2011
- 30↑
Satoer D, Visch-Brink E, Dirven C, Vincent A: Glioma surgery in eloquent areas: can we preserve cognition? Acta Neurochir (Wien) 158:35–50, 2016
- 31↑
Satoer D, Visch-Brink E, Smits M, Kloet A, Looman C, Dirven C, et al.: Long-term evaluation of cognition after glioma surgery in eloquent areas. J Neurooncol 116:153–160, 2014
- 32↑
Schmand B, Groenink SC, van den Dungen M: [Letter fluency: psychometric properties and Dutch normative data.] Tijdschr Gerontol Geriatr 39:64–76, 2008 (Dutch)
- 33↑
Spinhoven P, Ormel J, Sloekers PP, Kempen GI, Speckens AE, Van Hemert AM: A validation study of the Hospital Anxiety and Depression Scale (HADS) in different groups of Dutch subjects. Psychol Med 27:363–370, 1997
- 34↑
Vilasboas T, Herbet G, Duffau H: Challenging the myth of right nondominant hemisphere: lessons from corticosubcortical stimulation mapping in awake surgery and surgical implications. World Neurosurg 103:449–456, 2017
- 35↑
Xia L, Fang C, Chen G, Sun C: Relationship between the extent of resection and the survival of patients with low-grade gliomas: a systematic review and meta-analysis. BMC Cancer 18:48, 2018
- 36↑
Yushkevich PA, Gerig G: ITK-SNAP: an interactive medical image segmentation tool to meet the need for expert-guided segmentation of complex medical images. IEEE Pulse 8:54–57, 2017
- 37↑
Zigmond AS, Snaith RP: The Hospital Anxiety and Depression Scale. Acta Psychiatr Scand 67:361–370, 1983
