Supplementary materials proefschrift Eva van Grinsven
3 Appendices CHAPTER 2 SEARCH STRATEGY Search carried out on January 4th 2021 PICO process: Patient/population: Adult patients with brain metastases Intervention: Whole-brain radiotherapy (WBRT) and/or stereotactic radiosurgery (SRS) Comparison/control: Not applicable Outcome: Objective cognitive functioning Pubmed (“Brain Neoplasms/secondary”[Mesh] OR brain metastas*[Title/Abstract] OR metastatic brain*[Title/Abstract]) AND (Radiotherapy [MESH:NoExp] OR radiotherapy [Subheading] OR Radiosurgery [MESH] OR radiotherap*[Title/Abstract] OR radiation therap*[Title/Abstract] OR brain irradiation [Title/Abstract] OR radiosurg*[Title/Abstract] OR stereotactic radio* [Title/Abstract] OR SRS [Title/Abstract] OR SRT [Title/Abstract] OR gamma Knife [Title/Abstract] OR GKRS [Title/Abstract] OR GKS[Title/Abstract] OR GK[Title/ Abstract] OR whole brain radio* [Title/Abstract] OR WBRT [Title/Abstract]) AND (Neurocognitive Disorders [Mesh:NoExp] OR Cognition Disorders [Mesh:NoExp] OR Cognitive Dysfunction[Mesh] OR cognition [mesh] OR neuropsychology [MESH] OR Neuropsychological Tests [mesh] OR Memory [Mesh] OR Verbal learning [Mesh] OR Learning [Mesh] OR Psychomotor Performance [Mesh] OR Problem Solving [Mesh] OR Executive Function [Mesh] OR Mild Cognitive Impairment [Title/Abstract] OR Cogniti*[Title/Abstract] OR Neuropsychology* [Title/Abstract] OR Neurocogniti* [Title/Abstract] OR NCF [Title/Abstract] OR Memory [Title/Abstract] OR Verbal Learning [Title/Abstract] OR Learning [Title/Abstract] OR Psychomotor Performance [Title/Abstract] OR Problem Solving [Title/Abstract] OR Executive Function [Title/ Abstract] OR Digit Span [Title/Abstract] OR Digit Symbol [Title/Abstract] OR WAIS [Title/Abstract] OR Wechsler* [Title/Abstract] OR WMS [Title/Abstract] OR Trail Making Test [Title/Abstract] OR TMT [Title/Abstract] OR Hopkins Verbal Learning Test [Title/Abstract] OR HVLT [Title/Abstract] OR Hopkins Verbal Learning Test Revised [Title/Abstract] OR Rey Auditory Verbal Learning Test [Title/Abstract] OR RAVLT [Title/Abstract]OR Pegboard [Title/Abstract] OR Controlled Oral Word Association*
4 Supplementary materials [Title/Abstract] OR COWA [Title/Abstract] OR COWAT [Title/Abstract] OR Verbal Fluency [Title/Abstract] OR Semantic Fluency [Title/Abstract] OR Phonemic Fluency [Title/Abstract]) Embase (‘brain metastasis’/exp OR ‘brain metastas*’:ti,ab,kw OR ‘metastatic brain*’:ti,ab,kw) AND (‘radiotherapy’/de OR ‘radiosurgery’/de OR ‘skull irradiation’/exp OR ‘brain radiation’/ exp OR ‘stereotactic radiosurgery’/exp OR ‘stereotactic radiotherapy’/exp OR ‘gamma knife radiosurgery’/exp OR ‘radiotherap*’:ti,ab,kw OR ‘radiation therap*’:ti,ab,kw OR ‘brain irradiation’:ti,ab,kw OR ‘radiosurg*’:ti,ab,kw OR ‘stereotactic radio*’:ti,ab,kw OR ‘srs’:ti,ab,kw OR ‘srt’:ti,ab,kw OR ‘gamma knife’:ti,ab,kw OR ‘gkrs’:ti,ab,kw OR ‘gks’:ti,ab,kw OR ‘gk’:ti,ab,kw OR ‘whole brain radio*’:ti,ab,kw OR ‘wbrt’:ti,ab,kw) AND (‘disorders of higher cerebral function’/de OR ‘cognitive defect’/exp OR ‘cognition’/de OR ‘attention’/exp OR ‘cognitive reserve’/exp OR ‘executive function’/exp OR ‘verbal learning’/exp OR ‘mental capacity’/exp OR ‘social cognition’/exp OR ‘neuropsychology’/ exp OR ‘cognitive function test’/exp OR ‘neuropsychological test’/exp OR ‘memory’/ exp OR ‘memory disorder’/de OR ‘amnesia’/exp OR ‘attention disturbance’/exp OR ‘psychomotor performance’/de OR ‘mild cognitive impairment’:ti,ab,kw OR ‘cogniti*’:ti,ab,kw OR ‘neuropsychology*’:ti,ab,kw OR ‘neurocogniti*’:ti,ab,kw OR ‘ncf’:ti,ab,kw OR ‘attention’:ti,ab,kw OR ‘memory’:ti,ab,kw OR ‘verbal learning’:ti,ab,kw OR ‘learning’:ti,ab,kw OR ‘psychomotor performance’:ti,ab,kw OR ‘problem solving’:ti,ab,kw OR ‘executive function’:ti,ab,kw OR ‘digit span’:ti,ab,kw OR ‘digit symbol’:ti,ab,kw OR ‘wais’:ti,ab,kw OR ‘wechsler*’:ti,ab,kw OR ‘wms’:ti,ab,kw OR ‘trail making test’:ti,ab,kw OR ‘tmt’:ti,ab,kw OR ‘hopkins verbal learning test’:ti,ab,kw OR ‘hvlt’:ti,ab,kw OR ‘hopkins verbal learning test revised’:ti,ab,kw OR ‘rey auditory verbal learning test’:ti,ab,kw OR ‘ravlt’:ti,ab,kw OR ‘pegboard’:ti,ab,kw OR ‘controlled oral word association*’:ti,ab,kw OR ‘cowa’:ti,ab,kw OR ‘cowat’:ti,ab,kw OR ‘verbal fluency’:ti,ab,kw OR ‘semantic fluency’:ti,ab,kw OR ‘phonemic fluency’:ti,ab,kw) AND [embase]/lim AND (‘article’/it OR ‘article in press’/it) A
5 Appendices DIVISION OF NEUROPSYCHOLOGICAL TESTS Cognitive construct Neuropsychological test(s) Attention Digit Span (WAIS) Executive function Concept Shifting Test Key-search test (BADS) TMT B Fine motor coordination Grooved Pegboard Information processing speed Digit Symbol (WAIS) LDST SCWT TMT A Learning and memory EBPM HVLT(-R) ISLT List learning task (RBANS) OCLT RAVLT TBPM Verbal fluency COWA Phonemic and semantic fluency Semantic fluency Visual memory and visuoconstruction ROCFT Abbreviations: COWA, Controlled Oral Word Association Test; EBPM, event-based prospective memory; HVLT(-R), Hopkins Verbal Learning Test(- Revised); ISLT, International Shopping List Test; LDST, Letter Digit Substitution Test; MMSE, Mini-Mental Status Examination; OCLT, One Card Learning Test; RAVLT, Rey Auditory Verbal Learning Test; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status; ROCFT, Rey Osterrieth Complex Figure Test; SCWT, Stroop Color-Word Test; TBPM, time-based prospective memory, TMT, Trail Making Test; WAIS, Wechsler Adult Intelligence Scale.
6 Supplementary materials FOREST PLOTS Whole-brain radiotherapy Short-term follow-up Learning and memory Executive function Verbal fluency Information processing speed A
7 Appendices Mid-term follow-up Learning and memory Executive function Verbal fluency Information processing speed
8 Supplementary materials Long-term follow-up Learning and memory Executive function Verbal fluency Stereotactic radiosurgery Short-term follow-up Learning and memory Executive function A
9 Appendices Verbal fluency Information processing speed Fine motor coordination Attention Visual memory and visuoconstruction
10 Supplementary materials Mid-term follow-up Learning and memory Executive function Verbal fluency Information processing speed Fine motor coordination A
11 Appendices Attention Visual memory and visuoconstruction Long-term follow-up Learning and memory Executive function Verbal fluency Information processing speed
12 Supplementary materials Fine motor coordination A
13 Appendices CHAPTER 3 SUPPLEMENTARY METHODS Study set-up COIMBRA and APRICOT Study procedures The data was prospectively collected from the Cohort for patient-reported Outcomes, Imaging and trial inclusion in Metastatic BRAin disease (COIMBRA, NCT05267158) and from the Assessing and Predicting Radiation Influence on Cognitive Outcome using the cerebrovascular stress Test (APRICOT) study. COIMBRA candidates are minimally required to give consent for the collection and use of their clinical data in order to participate in the study. In addition, patients can decide to give consent for 1) filling in QoL questionnaires before radiotherapy, one month after, and subsequently every three months after radiotherapy, 2) undergoing NCAs before, three months and at least eleven months after radiotherapy, 3) undergoing additional MRIs, 3) future randomization for clinical trials, and 4) sharing of anonymized data with third parties. If patients gave consent for the COIMBRA study (including QoL-questionnaires) and fit the APRICOT in- and exclusion criteria (see below), they are subsequently asked to participate in the APRICOT study. For the APRICOT study NCAs are also performed before, three months and at least eleven months after radiotherapy. Additionally, extra MRI scans focusing on cerebrovascular functioning before and three months after radiotherapy are performed (see Supplementary Figure 1).
14 Supplementary materials Supplementary Figure 1. Study set-up of the COIMBRA and APRICOT studies. Blue tiles are only performed for the COIMBRA study, yellow tiles only for the APRICOT study and green tiles overlap between the studies. Dashed lines indicate patients can choose whether to participate in these study procedures. APRICOT in- and exclusion criteria For the APRICOT study patient additionally had to have an expected survival of ≥ 5 months (as determined by the Graded Prognosis Assessment score). Patients were not eligible if they had standard contraindications for 3T MRI scanning and/ or use of the RespirAct RA-MRTM MRI UNIT, had medical contraindications to limited hypercapnia, were unwilling/unable to cooperate with breathing maneuvers or keeping still, were noncompliant with prescribed anti-seizure medication, had severe neurological or psychiatric diseases not related to the primary malignancy or BMs, had a history of cerebrovascular disease, used >4 mg dexamethasone per day non-prophylactically, had cardiovascular disease, pulmonary disease, concurrent severe or uncontrolled medical disease, history of bleomycin treatment, a body weight <30 kg or >100 kg or were pregnant. Semi-structured interview Physical complaints (yes/no) • Headache (e.g. uni- or bilateral pain) • Motor problems (e.g. difficulties using arm(s)/leg(s)) • Sensory problems (e.g. different experience of touch, different sensation of warm/cold) • Visual problems (e.g. not being able to see part of the visual field, seeing blurry) A
15 Appendices • Coordination problems (e.g. difficulty fluently reaching/grasping for objects) • Nausea • Fatigue (e.g. having less energy to perform everyday tasks, needing to sleep more) Cognitive complaints Do you currently experience problems with…. • Memory o Remembering new names/faces o Remembering recent events or information o Remembering future plans o Remembering information from long ago • Orientation o Knowing the date, day and time o Knowing where you are o Knowing who the people around you are • Attention & executive functioning o Maintaining attention to things within/outside one’s own interest o Doing two things at the same time • Processing speed o Difficulty following conversations o Speed and structuring one’s own thoughts • Language o Difficulty with speaking (dysarthria) o Difficulty finding words (aphasia) o Difficulty understanding spoken or written words o Difficulty reading o Difficulty writing, making grammar/spelling mistakes • Emotional processing o Do you react to situations in a way that is appropriate in that situation? o Do you recognize your own emotional reaction to situations/person?
16 Supplementary materials Cognitive test battery Supplementary Table 1. Neuropsychological tests per neurocognitive domain. Light grey items are tests advised by the ICCTF and dark grey items tests that have often been performed in previous research. All grey items are part of the core battery. Premorbid intelligence Ravens Advanced Progressive Matrices39 Attention Wechsler Adult Intelligence Scale (WAIS-IV) Digit Span Forward Score 40 Trail Making Test (TMT) Switching ratio B/A 41,42 Stroop/Delis Kaplan Executive Function System (DKEFS) Switching ratio IV vs III 43 Executive functioning WAIS-IV Digit Span Backward Score 40 Letter fluency (3 letters) 44,45 Stroop D’KEFS Inhibition ratio III vs I 43 Memory Hopkins Verbal Learning Test – Revised (HVLT-R) Immediate, delayed, and recognition 46,47 Rey-Osterieth Complex Figure Test (ROCFT) Delayed copy 48 Visual Association Test (VAT) long form Immediate and delayed 49 Semantic fluency 50 Processing speed TMT A 41 Stroop/DKEFS naming [I] and reading [II] 43 Psychomotor speed Lafayette Grooved Pegboard Dominant and non-dominant hand 51 Visuospatial functioning ROCFT direct copy 48 Hooper Visual Organization Test (HVOT) Fragmented 52 Social cognition Facial Expressions of Emotion – Stimuli and Tests (FEEST) Total score 53 Language HVOT non-fragmented [clinical interpretation] 52,54 A
17 Appendices SUPPLEMENTARY RESULTS Baseline characteristics COIMBRA & APRICOT As the study population of two prospective studies was combined, we first compared patient-related and clinical characteristics of the COIMBRA patients with the APRICOT patients. Differences were assessed using chi-square tests for categorical data and t-tests or Mann-Whitney U-test for continuous data, depending on the distribution of the tested variable. In total, 58 patients underwent cognitive testing for either the COIMBRA (n = 32) or the APRICOT (n = 26) study between November 2020 and January 2023 (Supplementary Table 2). The APRICOT patients had a higher KPS than COIMBRA patients (p = .020), but did not differ regarding any of the other clinical characteristics. Most patients were scheduled to receive SRS (52/58), with 5/58 scheduled for WBRT and one patient for WBRT+SRS. Supplementary Table 2. Baseline characteristics in patient group. Clinical characteristic Total (n = 58) Age, years, median (IQR) 66.0 (57.8 – 72.3) Sex (male), n (%) 31 (53.4) Educational levela, n (%) 3 3 (5.2) 4 10 (17.2) 5 23 (39.7) 6 14 (24.1) 7 8 (13.8) Ravens matrices39 percentile score, median (IQR) 62.5 (37.5 – 84.4) Handednessb, n (%) Left 7 (12.1) Right 48 (82.8) Ambidextrous 3 (5.2) KPS, median (IQR)c 80 (80-90) No. of BMs, n (%) 1 16 (27.6) 2-4 25 (43.1) 5-10 10 (17.2) >10 7 (12.1) Hemisphere involvement BMs, n (%) Left 14 (24.1) Right 12 (20.7)
18 Supplementary materials Supplementary Table 2. Baseline characteristics in patient group. (continued) Clinical characteristic Total (n = 58) Bilateral 32 (55.2) Lobe involvement, n (%) Frontal 34 (58.6) Temporal 12 (20.7) Occipital 21 (36.2) Parietal 26 (44.8) Cerebellum 27 (46.6) Brainstem 3 (5.2) Primary tumor origin, n (%) Lung cancer 29 (50.0) Melanoma 14 (24.1) Breast Cancer 3 (5.2) Renal cell carcinoma 3 (5.2) Other 9 (15.5) Extracranial metastases, n (%) 34 (58.6) BMs as first symptom of cancer diagnosis, n (%) 18 (31.0) Previous brain RT, n (%) 9 (15.5) Previous BMs resection, n (%) 15 (25.9) Previous immuno-/chemotherapy, n (%) 43 (74.1) Dexamethasone use prior to RT, mg/day, median (IQR) 0.3 (0.0 – 4.0) Symptomatic BMs at diagnosis, n (%) 36 (62.1) Epilepsy d 10 (27.8) Motor d 11 (30.6) Sensory d 3 (8.3) Balance d 9 (25.0) Language d 1 (2.8) Visual d 8 (22.2) Cognitive d 7 (19.4) Headache d 12 (33.3) Other d 8 (22.2) aAccording to Verhage classification19, bself-reported, cn = 53, dPercentage of patients with symptomatic BMs at diagnosis. A
19 Appendices Subjective cognitive complaints Supplementary Table 3. Overview of reported subjective symptoms during the semistructured interview prior to NCA divided into physical and cognitive symptoms. Physical symptoms n (%) Cognitive symptoms n (%) Headache 12 (20.7) Memory 22 (37.9) Motor 22 (37.9) Orientation 3 (5.2) Coordination 10 (17.2) Attention & executive functioning 22 (37.9) Sensory 13 (22.4) Processing speed 14 (24.1) Visual 19 (32.8) Language 15 (25.9) Epilepsy 7 (12.1) Emotional processing 11 (19.0) Nausea 7 (12.1) Fatigue 36 (62.1) Neurocognitive functioning Supplementary Table 4. Cognitive test scores showing the group mean Z-scores for all neurocognitive (sub)tests and overall neurocognitive domain scores. Cognitive test or domain Mean Z score (SD) Range Attention (n = 55) 0.09 (0.72) -2.22 - 1.70 DS – Forward (n = 58) 0.17 (1.06) -2.00 - 2.67 TMT - Switching (n = 54)a 0.48 (0.88) -2.45 - 1.85 STROOP - Switching (n = 50)a -0.29 (1.11) -3.00 - 2.33 Executive function (n = 57) -0.13 (0.90) -2.33- 1.90 DS – Backward (n = 58) -0.18 (1.26) -2.67 - 3.00 Letter fluency (n = 56) -0.40 (1.07) -3.00 - 2.90 STROOP Inhibition (n = 51)a 0.23 (0.97) -2.67 - 2.00 Memory (n = 57)a -0.58 (0.97) -3.81- 0.56 HVLT-R - Immediate recall (n = 56)a -1.01 (1.27) -5.00 - 0.84 HVLT-R - Delayed recall (n = 56)a -0.84 (1.31) -4.33 - 0.82 HVLT-R - Delayed recognition (n = 56)a -1.34 (1.91) -8.25 - 0.81 ROCFT - Delayed Recall (n = 56) 0.16 (1.03) -2.27 - 2.19 VAT - Immediate Recall (n = 45)a -0.29 (1.11) -8.50 - 1.00 VAT - Delayed Recall (n = 43)a 0.17 (1.46) -6.67 - 0.67 Semantic fluency (2min) (n = 55)a -0.39 (0.85) -1.95 - 2.92 Processing speed (n = 57)a -0.63 (1.13) -5.02 – 0.59 TMT - Part A (n = 56)a -0.52 (1.21) -5.02 - 1.04 STROOP - Naming (n = 53)a -0.57 (1.08) -3.00 - 1.00
20 Supplementary materials Supplementary Table 4. Cognitive test scores showing the group mean Z-scores for all neurocognitive (sub)tests and overall neurocognitive domain scores. (continued) Cognitive test or domain Mean Z score (SD) Range STROOP - Reading (n = 53)a -0.38 (1.12) -3.00 - 1.00 Psychomotor speed (n = 56)a -1.45 (2.59) -17.70 – 0.71 GP - Dominant hand (n = 55)a -1.59 (3.09) -18.99 - 1.21 GP - Non-dominant hand (n = 55)a -1.28 (2.43) -15.35 - 1.40 Visuospatial functioning (n = 57)a 0.61 (0.68) -1.83 - 1.60 ROCFT - Direct Copy (n = 56)a 0.17 (1.14) -4.18 - 1.24 HVOT – Fragmented (n = 57)a 1.03 (0.62) -1.23 - 2.03 FEEST total (n = 34) -0.95 (1.10) -3.71 - 1.26 Bold values indicate scores that significantly differed from the normative population (i.e. Z = 0), with positive values indicating better and negative values indicating worse performance. Corrected p-value < .0025 for individual (sub)tests and corrected p-value < .008 for overall neurocognitive domain scores. a indicates domain or test-scores which were not-normally distributed and thus assessed using Wilcoxon signed-rank tests. All other measures were tested using one-sample t-tests. Abbreviations: DS, digit span; FEEST, Facial Expressions of Emotion – Stimuli and Tests; GP, Grooved Pegboard; HVLT-R, Hopkins Verbal Learning Test – Revised; HVOT, Hooper Visual Organization Test; ROCFT, Rey Osterieth Complex Figure Test; TMT, trail making test; VAT, Visual Association Test. Subjective and objective neurocognitive functioning Chi-square tests indicated the presence of subjective complaints regarding a cognitive domain did not discriminate between patients with and without objective impairments in that same domain. However, more patients with an attention impairment reported emotional complaints during the semi-structured interview (44%) compared to patients without an attention impairment (15%; p=.037, φ=.276). Additionally, subjective language complaints were more frequently reported by patients with a social cognition impairment (50%) compared to patients without a social cognition impairment (13%; p=.036, φ=.411). When evaluating change in subjective cognitive performance using the continuous VAS difference score, a Mann-Whitney U test indicated patients with visuospatial impairments reported a larger decline in perception from premorbid functioning on the VAS (M=-34.4, SD=32.9) than those without visuospatial impairments (M=-10.4, SD=18.5, p=.045). No differences were found between patients with or without cognitive impairments and VAS difference scores for any of the other tested cognitive domains. As psychomotor performance could be influenced by sensory and/or motor problems additional analyses were performed. Patients with sensory problems were A
21 Appendices more likely to have psychomotor speed impairment (67%) than patients without sensory problems (32%, p = .028). There was no differences between the groups with or without motor problems regarding the proportion of psychomotor speed impairments As half of the patients reporting sensory problems, reported those sensory problems to be due to neuropathy. Accordingly, patients were divided into three groups: 1) patients without sensory problems, 2) patients with sensory problems unrelated to neuropathy, and 3) patients with sensory problems due to neuropathy. There was no difference in the number of patients with impaired psychomotor speed performance between patients without sensory problems, neuropathy-unrelated sensory problems and neuropathy-related sensory problems (p = .084). There was no relationship between stress and cognitive deficits within any of the domains (all correlations p > .05). Feasibility of comprehensive test battery The completion rate of the comprehensive NCA battery (all 22 (sub)tests) versus the core test battery (11 subtests) was 52% and 91%, respectively. On average 91% of the 23 subtests of the comprehensive NCA battery were completed, while 52% of the patients completed all 22 (sub)tests (Supplementary Figure 2). Of the core battery, 96% of the tests was completed on average, while 91% of the patients completed the entire battery. The Facial Expressions of Emotion: Stimuli and Tests (FEEST) and Visual Association Test (VAT) were most often not completed, mostly due to time constraints as the brain radiotherapy was planned directly after the NCA (46% and 67% of missing cases, respectively). Of those patients finishing the full NCA (n = 29), median duration was 87 minutes (range 65-126 minutes) with 72% of all patients staying within intended the 90 minutes duration. Cognitive fatigue half-way and at the end of the full NCA was assessed using the Digit Span Forward. At the half-way mark 12/56 (21%) patients and at the end of the NCA 10/44 (23%) showed signs of fatigue (Supplementary Table 5). 6/44 patients (14%) showed signs of cognitive fatigue at both tested time points. Two patients were not able to complete the Digit Span Forward at the half-way mark due to self-reported fatigue. Additionally, nine patients could not finish the Digit Span Forward at the end of the NCA due to time constraints with other clinical appointments and two due to self-reported fatigue. To assess the impact of cognitive fatigue on the cognitive performance, MannWhitney U-tests were performed to compare the mean z-scores per domain
22 Supplementary materials across patients with and without cognitive fatigue (Supplementary Table 6). Patient with cognitive fatigue half-way through the NCA had an overall worse score regarding processing speed than patients without cognitive fatigue at the same time point. Performance on other cognitive domains did not differ. Additionally, to assess whether there was a difference in the amount of patients with a cognitive impairment between patients with and without cognitive fatigue chi-square tests were performed (Supplementary Table 7). No significant differences were found for any of the cognitive domains. Supplementary Figure 2. Reasons for not completing cognitive tasks. Light grey colored boxes indicate tasks that are part of the core battery. Icons indicate the reasons for not completing a task as indicated by the legend in the upper right corner. Abbreviations: FEEST, Facial Expressions of Emotion – Stimuli and Tests; HVLT-R, Hopkins Verbal Learning Test – Revised; HVOT, Hooper Visual Organization Test; ROCT, Rey Osterieth Complex Figure Test; TMT, trail making test; VAT, Visual Association Test. A
23 Appendices Supplementary Table 5. Number and percentage of patient with a difference in maximum span half-way and at the end of the NCA. Difference in maximum span Half-way the NCA End of the NCA 0 44 (78.6%) 34 (77.3%) -1 11 (19.6%) 8 (18.2%) -2 1 (1.8%) 2 (4.5%) Supplementary Table 6. Comparison of mean (SD) z-scores per cognitive domain between patients with and without fatigue either half-way or at the end of the NCA. Fatigue half-way through NCA Fatigue at the end of NCA Yes (n =12) No (n = 44) p-value Yes (n =10) No (n = 34) p-value Attention -0.04 (0.85) 0.13 (0.68) .775 0.02 (0.76) -0.02 (0.69) .634 Executive function -0.10 (0.69) -0.11 (0.94) .795 0.13 (0.92) -0.26 (0.90) .157 Memory -0.78 (1.34) -0.48 (0.80) .889 -0.80 (1.44) -0.46 (0.84) .889 Processing speed -1.35 (1.43) -0.41 (0.98) .006 -0.38 (0.88) -0.28 (0.77) .634 Psychomotor speed -0.83 (1.05) -1.60 (2.86) .542 -0.24 (0.69) -1.52 (3.06) .053 Visuospatial functioning 0.70 (0.60) 0.58 (0.71) .563 0.80 (0.60) 0.59 (0.73) .334 Social cognition -1.22 (1.29) -0.86 (1.05) .611 -0.98 (1.14) -0.89 (1.02) .831 Bold values indicate statistically significant difference between patients with and without fatigue, corrected p-value < .007.
24 Supplementary materials Supplementary Table 7. Comparison of the number (%) of patients with cognitive impairment between patients with and without fatigue either half-way or at the end of the NCA. Fatigue half-way through NCA Fatigue at the end of NCA Yes (n =12) No (n = 44) p-value Yes (n =10) No (n = 34) p-value Attention 2 (16.7) 7 (15.9) .949 2 (20.0) 7 (20.6) .968 Executive function 3 (25.0) 9 (20.5) .734 2 (20.0) 9 (26.5) .678 Memory 7 (58.3) 22 (50.0) .609 5 (50.0) 18 (52.9) .870 Processing speed 7 (58.3) 11 (25.0) .028 3 (30.0) 7 (20.6) .532 Psychomotor speed 3 (27.3) 18 (40.9) .405 1 (10.0) 12 (35.3) .123 Visuospatial functioning 0 (0.0) 5 (11.4) .221 0 (0.0) 3 (8.8) .331 Social cognition 3 (50.0) 7 (35.0) .508 2 (28.6) 6 (37.5) .679 Corrected p-value < .007. Comparison between comprehensive versus core battery A severe impairment (-2.0SD) in at least one task within one domain was found for 71% for the comprehensive battery and 59% for the core battery. A more subtle deficit (-1.5SD) for at least one domain was found for 79% on the comprehensive and 71% on the core battery. When using the ICCTF criterion for impairment (≤-1.5SD on ≥2 test or ≤-2.0SD on ≥1 test) 76% of all patients were classified as cognitively impaired when all 10 NCA tests were considered and 60% when only the core battery was taken into account. We additionally examined the effects of using the ICCTF cognitive domain categorization18 on differences between the comprehensive and the core battery. Based on the ICCTF, both the TMT switching ratio (B/A) and STROOP switching ratio (IV vs III) were grouped under the domain of executive function instead of attention. Upon examination, results indicated that all patients with deficits on either the TMT switching ratio (B/A) and STROOP switching ratio (IV vs III) did not show deficits in any of the other executive function tests. Therefore, according to the ICCTF cognitive domain grouping, these patients were categorized as having executive function impairments rather than attention deficits. This indicates the number of cognitive domains found to be impaired with either the comprehensive or the core battery remained the same, regardless of the domain classification that was used. Thus, A
25 Appendices the results as shown in comparing the comprehensive and the core battery are unaffected by the cognitive domain classification. Supplementary Figure 3. Comparison of patient classification based on number of cognitive impairments when using the core test battery versus the comprehensive test battery. Each icon indicates one patient where the color of the icon indicates the number of cognitively impaired domains per individual when evaluated using the comprehensive cognitive test battery. Strainers indicate the number of impaired domains needed to be caught in the strainer when solely using the core test battery. Patients shown on the right of the figure are those that were ‘misclassified’ if only the core test battery was taken into account.
26 Supplementary materials CHAPTER 4 SUPPLEMENTARY METHODS Neurocognitive assessment Supplementary Table 1. Neuropsychological tests per neurocognitive domain. Premorbid intelligence Ravens Advanced Progressive Matrices29 Attention Wechsler Adult Intelligence Scale (WAIS-IV) Digit Span Forward Score 52 Trail Making Test (TMT) Switching ratio B/A 53,54 Stroop/Delis Kaplan Executive Function System (DKEFS) Switching ratio IV vs III 55 Executive functioning WAIS-IV Digit Span Backward Score 52 Letter fluency (3 letters) 56,57 Stroop D’KEFS Inhibition ratio III vs I 55 Memory Hopkins Verbal Learning Test – Revised (HVLT-R) Immediate, delayed, and recognition 58,59 Rey-Osterieth Complex Figure Test (ROCFT) Delayed copy 60 Visual Association Test (VAT) long form Immediate and delayed 61 Semantic fluency 62 Processing speed TMT A 53 Stroop/DKEFS naming [I] and reading [II] 55 Psychomotor speed Lafayette Grooved Pegboard Dominant and non-dominant hand 63 Visuospatial functioning ROCFT direct copy 60 Hooper Visual Organization Test (HVOT) Fragmented 64 Social cognition Facial Expressions of Emotion – Stimuli and Tests (FEEST) Total score 65 Language HVOT non-fragmented [clinical interpretation] 64,66 A
27 Appendices Reliable Change Index Normative data Supplementary Table 2. Normative data used for the calculation of the RCI. Cognitive test N Male/Female ratio Mean age (SD) Education in years Retest interval Digit Span [forward, backward]67 48 32/16 15.9 N.A. 60 days FEEST [negative emotions] 16 8/8 24 (4.3) N.A. 6.19 - 7.25 days Grooved pegboard [dominant and non-dominant hand]68 605 605/0 39.5 (8.7) 16.4 (2.3) N.A. HVLT-R [immediate, delayed, and recognition]58 541 200/341 48.1 (17.3) 13.8 (2.3) N.A. HVOT [derived VOT]69 N.A. N.A. N.A. N.A. N.A. Letter fluency70 120 0/120 N.A. N.A. 1 year ROCFT [direct and delayed copy]68 478 478/0 42.2 (8.6) 16.4 (2.3) N.A. Semantic fluency [60 seconds]71 188 71/117 75 13.1 (3.7) N.A. Stroop DKEFS [I-IV]70 45 0/45 18-70 N.A. 6 months TMT [A, B/A]72 355 166/189 55.0 (18.3) 11.9 (4.6) N.A. VAT Long Form [immediate recall]70 120 0/120 N.A. N.A. 1 year Abbreviations: FEEST, Facial Expressions of Emotion – Stimuli and Tests; HVLT-R, Hopkins Verbal Learning Test – Revised; HVOT, Hooper Visual Organization Test; N.A., not available; ROCFT, Rey Osterieth Complex Figure Test; TMT, trail making test; VAT, Visual Association Test.
28 Supplementary materials SUPPLEMENTARY RESULTS Baseline subjective cognitive performance Supplementary Figure 1. Subjective cognitive performance comparing pre-radiotherapy to the pre-cancer performance. Note: stable performance (±5), subtle improvement or decline (±6-25), substantial improvement or decline (±26-50) and extreme improvement or decline (±>50). Post-radiotherapy subjective cognitive performance Supplementary Table 3. Factors associated with 3 months post-radiotherapy changes in subjective cognitive performance. Variable Decline, n (%) Improvement, n (%) Mixed, n (%) Stable, n (%) p Age ≤65 6 (30) 6 (30) 4 (20) 4 (20) .248 ≥66 8 (50) 5 (31) 0 (0) 3 (19) KPS ≤80 7 (32) 6 (27) 3 (14) 6 (27) .386 ≥90 7 (50) 5 (36) 1 (7) 1 (7) Primary tumor Lung 6 (35) 6 (35) 3 (18) 2 (12) .465 Other 8 (42) 5 (26) 1 (5) 5 (26) Extracranial metastases A
29 Appendices Supplementary Table 3. Factors associated with 3 months post-radiotherapy changes in subjective cognitive performance. (continued) Variable Decline, n (%) Improvement, n (%) Mixed, n (%) Stable, n (%) p No 5 (33) 6 (40) 1 (7) 3 (20) .702 Yes 9 (43) 5 (24) 3 (14) 4 (19) No of BMs 1-4 3 (27) 4 (36) 2 (18) 2 (18) .694 ≥5 11 (44) 7 (28) 2 (8) 5 (20) Symptomatic BMs No 6 (29) 8 (38) 4 (19) 3 (14) .118 Yes 8 (53) 3 (20) 0 (0) 4 (27) Synchronous BMs diagnosis No 9 (38) 6 (25) 3 (13) 6 (25) .566 Yes 5 (42) 5 (42) 1 (8) 1 (8) Intracranial progression No 5 (22) 8 (35) 4 (17) 6 (26) .031 Yes 9 (69) 3 (23) 0 (0) 1 (8) Baseline cognitive impairment 0 domains 2 (29) 5 (57) 1 (14) 0 (0) .383 1 domain 3 (27) 3 (27) 4 (36) 4 (36) ≥2 domains 9 (50) 4 (22) 2 (11) 3 (17) Bold values indicate statistically significant differences. Objective neurocognitive functioning Supplementary Table 4. Cumulative percentage of patients with a Z-score below the threshold for all time points and each task and cognitive domain. Cognitive test or domain Baseline 3 months ≥11 months -1 -1.5 -2 -1 -1.5 -2 -1 -1.5 -2 Attention 44.4 19.4 19.4 52.8 30.6 27.8 35.7 14.3 14.3 DS – Forward 27.8 5.6 5.6 19.4 2.8 2.8 0.0 0.0 0.0 TMT - Switching 8.3 2.8 2.8 13.9 11.1 8.3 7.7 0.0 0.0 STROOP - Switching 20.6 14.7 14.7 25.0 19.4 19.4 30.8 15.4 15.4 Executive function 50.0 25.0 16.7 47.2 25.0 13.9 50.0 35.7 28.6 DS – Backward 33.3 13.9 11.1 36.1 8.3 8.3 21.4 7.1 7.1 Letter fluency 27.8 13.9 5.6 27.8 13.9 5.6 21.4 14.3 7.1
30 Supplementary materials Supplementary Table 4. Cumulative percentage of patients with a Z-score below the threshold for all time points and each task and cognitive domain. (continued) Cognitive test or domain Baseline 3 months ≥11 months -1 -1.5 -2 -1 -1.5 -2 -1 -1.5 -2 STROOP Inhibition 14.7 5.9 2.9 8.3 5.6 0.0 15.4 15.4 15.4 Learning & memory 72.2 47.2 36.1 94.4 77.8 52.8 78.6 71.4 42.9 HVLT-R - Immediate recall 38.9 30.6 22.2 55.6 41.7 27.8 64.3 42.9 21.4 HVLT-R - Delayed recall 44.4 22.2 16.7 47.2 30.6 22.2 50.0 35.7 21.4 HVLT-R - Delayed recognition 58.3 33.3 33.3 50.0 30.6 30.6 28.6 21.4 21.4 ROCFT - Delayed Recall 13.9 8.3 2.8 11.1 2.8 2.8 23.1 15.4 7.7 VAT - Immediate Recall 21.9 6.3 3.1 41.7 30.6 25.0 50.0 35.7 28.6 VAT - Delayed Recall 3.3 3.3 3.3 5.7 5.7 5.7 30.8 23.1 23.1 Semantic fluency (2min) 22.2 2.8 0.0 47.2 25.0 2.8 21.4 0.0 0.0 Processing speed 41.7 30.6 22.2 36.1 25.0 22.2 69.2 30.8 23.1 TMT - Part A 19.4 11.1 5.6 16.7 11.1 8.3 23.1 15.4 7.7 STROOP - Naming 25.0 16.7 13.9 30.6 22.2 22.2 46.2 7.7 7.7 STROOP - Reading 22.2 16.7 11.1 22.2 16.7 11.1 38.5 23.1 23.1 Psychomotor speed 44.4 36.1 27.8 41.7 38.9 30.6 53.8 38.5 15.4 GP - Dominant hand 41.7 33.3 19.4 30.6 27.8 19.4 46.2 30.8 15.4 GP - Non-dominant hand 33.3 25.0 16.7 33.3 30.6 22.2 33.3 16.7 8.3 Visuospatial functioning 33.3 16.7 16.7 19.4 5.6 5.6 15.4 7.7 7.7 ROCFT - Direct Copy 11.1 8.3 8.3 2.8 2.8 2.8 7.7 7.7 7.7 HVOT – Fragmented 25.0 11.1 11.1 19.4 5.6 5.6 7.7 0.0 0.0 FEEST total 48.0 32.0 16.0 38.7 26.7 16.7 42.9 42.9 14.3 Abbreviations: DS, digit span; FEEST, Facial Expressions of Emotion – Stimuli and Tests; GP, Grooved Pegboard; HVLT-R, Hopkins Verbal Learning Test – Revised; HVOT, Hooper Visual Organization Test; ROCFT, Rey Osterieth Complex Figure Test; TMT, trail making test; VAT, Visual Association Test. A
31 Appendices Supplementary Figure 2. Percentage of patients without a cognitive impairment on the domain-level as well as overall at each time point.
32 Supplementary materials Supplementary Table 5. Factors associated with 3 months post-radiotherapy changes in objective cognitive performance. Variable Decline, n (%) Improvement, n (%) Mixed, n (%) p Age ≤65 3 (15) 0 (0) 17 (85) .488 ≥66 3 (19) 1 (6) 12 (75) KPS ≤80 3 (14) 1 (5) 18 (82) .619 ≥90 3 (21) 0 (0) 11 (79) Primary tumor Lung 3 (18) 0 (0) 14 (82) .629 Other 3 (16) 1 (5) 15 (79) Extracranial metastases No 2 (13) 0 (0) 13 (87) .605 Yes 4 (19) 1 (5) 16 (76) No of BMs 1-4 1 (9) 1 (9) 9 (82) .244 ≥5 5 (20) 0 (0) 20 (80) Symptomatic BMs No 2 (13) 1 (7) 12 (80) .456 Yes 4 (19) 0 (0) 17 (81) Synchronous BMs diagnosis No 5 (21) 0 (0) 19 (79) .251 Yes 1 (8) 1 (8) 10 (83) Intracranial progression No 3 (23) 0 (0) 10 (77) .577 Yes 3 (13) 1 (4) 19 (83) Baseline cognitive impairment 0 domains 2 (29) 1 (14) 4 (57) .099 1 domain 3 (27) 0 (0) 8 (73) ≥2 domains 1 (6) 0 (0) 17 (94) A
33 Appendices CHAPTER 5 SUPPLEMENTARY METHODS In- and exclusion criteria Tumor patients Patients who had any prior cancer treatment (i.e. operation [except biopsy before resection], chemotherapy, radiotherapy) were not included in the retrospective cohort. Additional inclusion criteria for the current study were availability of preoperative neurocognitive assessment, fluent in Dutch and a pre-operative T2 FLAIR MRI with slice thickness ≤5 mm in order to maintain adequate quality for lesion segmentation. No criterion was set for the amount of completed neurocognitive tasks. The tumor location on the T2 FLAIR had to be representative of the tumor at the time of the neurocognitive assessment as visible on other imaging modalities as determined by the neurologist (TS). Glioma patients with a history of neurological or psychiatric disease potentially interfering with the current neurocognitive performance or imaging were excluded for the current study. Stroke patients Stroke patients were asked to participate if they were admitted to any of the following hospitals in the Netherlands between September 2015 and December 2019 in the Netherlands: Amsterdam University Medical Center (Amsterdam UMC), Radboud University Medical Center (Radboudumc) in Nijmegen, University Medical Center Groningen (UMCG), University Medical Center Utrecht (UMCU), Onze Lieve Vrouwe Gasthuis (OLVG), Maasziekenhuis Pantein, Rijnstate, Ommelander Ziekenhuis Groep, St. Antonius Ziekenhuis, and Diakonessenhuis. Patients were included for the current analyses if they were diagnosed with cerebral ischemic stroke made by a neurologist and fluent in Dutch. Exclusion criteria were 1) prior stroke based on imaging, 2) presence of another neurological, psychiatric or other diagnosis that may interfere with cognitive testing/imaging, 3) pre-existing cognitive decline (using a short informant-based questionnaire to asses change in cognitive functioning prior to stroke; cognitive decline was defined as a score >3.6 on the Dutch version of the Informant Questionnaire on Cognitive Decline in the Elderly [IQCODE]81). Neurocognitive assessment All subjects underwent a neurocognitive assessment that included measures of language, attention, executive functions, memory, visuoconstruction and psychomotor speed. The neurocognitive assessment was administered via the
34 Supplementary materials standardized instructions and scoring was performed according to the standardized scoring criteria. For tumor patients, the assessment was part of routine clinical care and the exact selection of tests was based on a standard protocol, which was tailored individually to patients’ complaints, tumor location and time constraints. Overall, most tumor patients completed the neurocognitive assessment within approximately two hours, including a break halfway through. For the stroke patients a standard battery of tests was used. Overall, most stroke patients completed the neurocognitive assessment within approximately one and a half hours. Supplementary Table 1 shows the neuropsychological tasks and corresponding scores that were available for both the tumor and stroke sample. For the lesion-symptom mapping analyses, we selected the Rey Auditory Verbal Learning Test (RAVLT, direct recall, delayed recall, and delayed recognition) and the verbal fluency test (Dutch versions of the Controlled Oral Word Association Test (COWAT) and Category Fluency (animal), both described in Bouma et al.82). The RAVLT is a verbal learning and memory test that taps into multiple partly dissociated aspects of memory. Specifically, it assesses the ability to learn new information, consolidate it, reproduce it and recognize it after a delay period. During the task 15 unrelated nouns are read aloud on five consecutive trials. Each learning trials is followed by a free immediate recall test in which participants are asked to name as many words as they can remember. After a 20-minute delay, the participant is asked to recall the words from the list presented during the learning trials. Additionally, a list of 30 words is presented and the participant must identify the 15 previously presented words. From the RAVLT, we used the total number of words remembered on the five learning trials (immediate recall), the number of words remembered after the delay period (delayed recall) and number of correctly identified words in the recognition trial (delayed recognition). The verbal fluency test is a short test of verbal functioning. More specifically, it assesses an individual’s ability to retrieve verbal information within restricted search parameters. The test is separated into a phonemic and semantic fluency part. In the phonemic fluency participants are given 60 seconds to generate as many unique words beginning with a single letter. Three trials, each with a different letter, are performed. In the semantic fluency test participants are given 60 seconds to generate unique words belonging to a certain semantic category, in this case animals. Both tasks require a complex interplay of a variety of cognitive functions including attention, vocabulary knowledge, retrieval of lexical and semantic knowledge and executive functions. The phonemic fluency test is thought to rely more heavily on executive control, while the semantic fluency test is more dependent on correct A
35 Appendices retrieval of semantic knowledge. For the letter fluency we used total number of correct words after three 60-seconds trials and for the semantic fluency the number of correct words within the category. Supplementary Table 1. Neuropsychological tasks and corresponding scores. Neuropsychological task Cognitive domain Raw score Boston Naming Testa Naming ability Total correct Letter fluencyb Verbal fluency Total amount of correct, unique words over three 60-seconds trials Semantic fluencyc Verbal fluency Total amount of correct, unique words within 60 seconds RAVLT - Dutch versiond Verbal memory Immediate recall (trial 1-5) Delayed recall score Delayed recognition score ROCFTe Visuoconstructive abilities Direct copy WAIS-III or WAIS-IV Digit Spanf Attention Span Forward span Working Memory Backward span Trail Making Testg Psychomotor speed Time to complete part A Switching Ratio score (B/A) Abbreviations: RAVLT, Rey Auditory Verbal Learning Test; ROCFT; Rey-Osterieth Complex Figure Test; WAIS, Wechsler Adult Intelligence Scale; aBoston Naming Task, Heesbeen, Van Loon-Vervoorn, 2001. bPhonologic Verbal Fluency Test (Lexical Fluency) (Harrison, Buxton, Husain, Wise, 2010; Schmand, Groenink, Van Den Dungen, 2008). cSemantic Verbal Fluency Test, Harrison et al, 2010 (Gerritsen et al., Nederlands Instituut van Psychologen, 2012). d15 Words Test (15WT) (Saan, Deelman, 1986). eBerry, Carpenter, 1992; Spreen, Strauss, 1998 fWechsler Adult Intelligence Scale Third Edition Digit Span (WAIS-III) (WAIS-III Administration and scoring manual, 1997), Wechsler Adult Intelligence Scale Fourth Edition Digit Span (WAIS-IV) (WAIS-IV-NL Technische handleiding, 2013).g Giovagnoli, Del Pesce, Mascheroni, Simoncelli, Laiacona, Capitani, 1996 (Drane, Yuspeh, Huthwaite, & Klinger, 2002)
36 Supplementary materials LESION VOLUME Supplementary Table 2. Spearman Rho correlation values and p values for the association between lesion volume and performance on cognitive tasks Glioma Stroke Total Boston Naming Test r = -.288 p < .001 r = -.099 p = .163 r = -.362 p < .001 Phonological fluency r = -.363 p < .001 r = -0.49 p = .339 r = -.126 p = .045 Semantic fluency r = -.341 p < .001 r = -.038 p =.336 r = -.095 p = .059 RAVLT – direct recall r = -.272 p < .001 r = -.021 p = .409 r = -.078 p = .088 RAVLT – delayed recall r = -.238 p < .001 r = .031 p = .366 r = -.101 p = .041 RAVLT – delayed recognition r = -.254 p < .001 r = -.143 p = .057 r = -.177 p = .001 ROCFT direct copy r = -.198 p = .005 r = -.135 p = .306 r = -.032 p < .310 WAIS digit span forward r = -.215 p = .001 r = -.137 p = .057 r = -.240 p < .001 WAIS digit span backward r = -.261 p < .001 r = -.100 p = .125 r = -.435 p < .001 TMT A r = -.276 p < .001 r = -.210 p = .009 r = -.119 p = .019 TMT B ratio r = -.180 p = .010 r = .082 p = .182 r = -.087 p = .068 Mean overall Z score r = -.415 p < .001 r = -.139 p = .051 r = -.172 p = .001 A
37 Appendices Supplementary Figure 1. Swarm chart of lesion volumes for the tumor and stroke group. Each dot represents the lesion volume of one patient with stroke patients represented by blue and tumor patients by red dots.
38 Supplementary materials Supplementary Table 3. Demographic and cognitive characteristics of the stroke group, subdivided by lesion volume. Stroke Lesion volume ≤3 cc >3 cc N 54 93 Sex (%) Male 39 (72.2) 60 (64.5) Female 15 (27.8) 33 (35.5) Mean age (SD) 57.3 57.5 Range 19-81 20-82 Lesion location (%) Left 27 (50.0) 39 (41.9) Right 20 (37.0) 46 (49.5) Bilateral 7 (13.0) 8 (8.6) Cognitive performance ≤-1.5 (%)* Boston Naming Test 2 (3.7) 0 (0.0) Digit Span Forward 5 (9.4) 10 (11.4) Digit Span Backward** 5 (9.4) 21 (23.9) RAVLT – Direct recall 5 (9.8) 12 (14.6) RAVLT – Delayed recall 9 (17.6) 13 (16.0) RAVLT – Delayed recognition 6 (11.8) 8 (9.9) ROCFT – Copy 11 (20.4) 11 (23.4) Semantic fluency** 1 (2.1) 13 (15.5) Letter fluency 6 (11.1) 10 (19.6) TMT A 2 (3.8) 8 (9.8) TMT B/A ratio 2 (4.0) 5 (6.2) * valid percentage is shown thus based on N of patients with a specific test score ** significant difference (p < .05) between groups A
39 Appendices COMBINED SVR-LSM We combined the data of both groups and performed the SVR-LSM analyses for each cognitive task using etiology (tumor or stroke) as a covariate on both the behavioral scores and lesion data. In areas where both groups have adequate coverage, and thus etiology can be used as covariate, this analysis allows investigation of the relation between lesion location and cognitive performance, irrespective of etiology. It is important to note that this analysis is still subject to differences in sample size between the groups and results can still be driven by one etiology simply because its sample size is significantly larger. Direct recall verbal memory (Supplementary Figure 2) When both groups were combined for the SVR-LSM analyses, lesions in the left ILF (most voxels with peak significance; 11.8%) were most strongly associated with performance on the RAVLT direct recall. Lesions in the left optic radiation (17.3%) and posterior segment (12.2%) were also associated with worse task performance. Additionally multiple grey areas were associated with direct recall performance, among which the left hippocampus (38.2%), the thalamus (15.5%), the parahippocampal gyrus (14.8%), the inferior temporal gyrus (14.4%) and the caudate nucleus (13.4%). When comparing these results to the etiology-specific lesion-symptom maps, largely overlapping voxels were found (Supplementary Figure 2 and Supplementary Table 5). Nevertheless, some brain areas were only significant in the combined maps (left inferior temporal gyrus) and others only in the etiology specific maps (left uncinate fasciculus, lingual gyrus).
40 Supplementary materials Supplementary Figure 2. SVR-LSM results for the RAVLT direct recall for both groups combined when etiology was added as covariate. Panel A shows the lesion overlap for this task when lesions from both groups are combined. The color bar indicates the number of patients with overlapping lesions. Panel B shows the voxels that were significantly associated with worse performance (yellow). The yellow color indicates the p-value for each voxel. Delayed recall verbal memory (Supplementary Figure 3) In the combined SVR-LSM analysis, lesions in the left ILF (most voxels with peak significance; 26.2%), optic radiation (29.2%), posterior segment (14.7%) and IFOF (12.3%) were strongly associated with task performance. Significant voxels extended into grey matter areas, including the left hippocampus (39.2%), parahippocampal gyrus (18.1 %), inferior temporal gyrus (12.3%). Significant areas largely overlapped with the areas found when the analysis was run for the tumor and stroke group separately (Supplementary Figure 3 and Supplementary Table 5). However, while lesions in the left putamen were highly associated with worse task performance in the stroke group, this area was not significant in the combined group SVR-LSM analysis. A
41 Appendices Supplementary Figure 3. SVR-LSM results for the RAVLT delayed recall for both groups combined when etiology was added as covariate. Panel A shows the lesion overlap for this task when lesions from both groups are combined. The color bar indicates the number of patients with overlapping lesions. Panel B shows the voxels that were significantly associated with worse performance (yellow). The yellow color indicates the p-value for each voxel. Delayed recognition verbal memory (Supplementary Figure 4) Multiple brain areas in of the left hemisphere were significantly associated with task performance in the combined group analysis. Significant grey matter brain areas included the middle temporal gyrus (most voxels with peak significance; 38.1%), the hippocampus (50.0%), the inferior temporal gyrus (32.9%), the parahippocampal gyrus (14.9%), the fusiform gyrus (11.6%) and the superior temporal gyrus (11.2%). Significant voxels extended into white matter areas like the ILF (58.9%), posterior segment (31.4%), optic radiation (31.1%) and the IFOF (14.5%). Although most brain areas overlapped with those found in the etiology specific lesion-symptom maps some brain areas were found in the combined group analysis that were not found when analyzing the groups separately, like the left superior temporal gyrus (Supplementary Figure 4 and Supplementary Table 5). Additionally, all right-sided brain areas that were involved in task performance in the stroke group (inferior and middle frontal gyrus), were not indicated by the combined group analysis.
42 Supplementary materials Supplementary Figure 4. SVR-LSM results for the RAVLT delayed recognition for both groups combined when etiology was added as covariate. Panel A shows the lesion overlap for this task when lesions from both groups are combined. The color bar indicates the number of patients with overlapping lesions. Panel B shows the voxels that were significantly associated with worse performance (yellow). The yellow color indicates the p-value for each voxel. Letter fluency (Supplementary Figure 5) Lesions within the left insula (most voxels with peak significance, 22.7%) were most strongly associated with worse letter fluency performance when the groups were combined. Additionally, the left putamen (35.1%), inferior frontal gyrus opercular (24.4%), caudate nucleus (15.9%) and long segment (12.1%) were significantly involved in the task performance. In general, there was good consensus with brain areas found in the etiology specific and combined lesion-symptom maps (Supplementary Figure 5 and Supplementary Table 5). A
43 Appendices Supplementary Figure 5. SVR-LSM results for the letter fluency for both groups combined when etiology was added as covariate. Panel A shows the lesion overlap for this task when lesions from both groups are combined. The color bar indicates the number of patients with overlapping lesions. Panel B shows the voxels that were significantly associated with worse performance (yellow). The yellow color indicates the p-value for each voxel. Semantic fluency (Supplementary Figure 6) The combined SVR-LSM analyses indicated lesions in the left corticospinal tract (most voxels with peak significance; 14.4%) to be most strongly related to worse semantic fluency performance. Significant voxels extended into grey matter areas, including the caudate nucleus (23.1%) and the precentral gyrus (21.5%). While at large the brain areas found in the etiology specific lesions-symptom maps were also found in the combined lesion-symptom maps, some differences were apparent. For example, the caudate nucleus was only significant when both groups were combined and the left superior dorsolateral frontal gyrus and rolandic operculum were only found in the separate tumor and stroke analyses, respectively. (Supplementary Figure 6 and Supplementary Table 5).
44 Supplementary materials Supplementary Figure 6. SVR-LSM results for the semantic fluency for both groups combined when etiology was added as covariate. Panel A shows the lesion overlap for this task when lesions from both groups are combined. The color bar indicates the number of patients with overlapping lesions. Panel B shows the voxels that were significantly associated with worse performance (yellow). The yellow color indicates the p-value for each voxel. A
www.ridderprint.nlRkJQdWJsaXNoZXIy MTk4NDMw