Martin Taphoorn

Background: It is unclear whether glioma subtypes differ in health-related quality of life (HRQoL) over time. Methods: We analyzed data from the CODAGLIO 2.0 database, comprising 16 randomized controlled trials with HRQoL assessed using the EORTC QLQ-C30 and QLQ-BN20 questionnaires. Differences in mean HRQoL at baseline (before postoperative treatments) and 3 months were compared across glioma subtypes. Linear mixed models (LMMs) were used to identify factors independently associated with HRQoL, and anchor-based minimally important differences determined clinical relevance. Results: A total of 4,301 patients had baseline HRQoL data; 75% completed follow-up. In stratified analyses, at 3 months, physical, social, role, and cognitive functioning were statistically and clinically significantly worse in glioblastoma than in astrocytoma or oligodendroglioma. In adjusted LMMs, compared with glioblastoma, patients with astrocytoma had worse physical (−2.78, 95% CI, −4.68 to −1.01) and cognitive functioning (−3.63, −6.18 to −1.05), more pain (2.74, 0.41-4.95), communication deficits (5.03, 2.44-7.66), and seizures (4.49, 2.55-6.72). Patients with oligodendroglioma had worse social functioning (−5.39, −10.57 to −0.11), communication deficits (7.40, 1.63-12.81), and seizures (8.61, 4.17-14.26). None of these differences exceeded clinical relevance thresholds. Performance status was the only variable associated with clinically meaningful HRQoL changes across most scales. Conclusions: At 3 months, patients with glioblastoma show clinically worse HRQoL than IDH-mutant glioma patients in unadjusted analyses. However, after adjusting for clinical variables, glioblastoma patients exhibit statistically but not clinically better HRQoL in several domains. Performance status emerges as the key factor independently associated with HRQoL and should play a central role in clinical decision-making.

Background: The EORTC QLQ-C30 Summary Score was developed to address analytical challenges arising from the multidimensional profile of the QLQ-C30. We compared the construct validity and responsiveness to change of the Summary Score with the Global Health/QoL scale and other individual QLQ-C30 scales to compare their performance as HRQoL outcome measures.Methods: We conducted a retrospective analysis of pooled data from six EORTC randomized clinical trials (n=3,246). Construct validity was assessed using cross-sectional known-group comparisons based on tumour stage, WHO performance status, and survival time. Responsiveness was evaluated using baseline and on-treatment assessments. Effect sizes were calculated using Cohen’s d, and relative validity (RV) was used to compare discriminatory performance across scales.Results: The Summary Score showed strong discriminative ability in known-group comparisons, with Cohen’s d effect sizes of 1.24 (tumour stage), 0.90 (WHO performance status), and 0.85 (survival time), consistently outperforming the Global Health/QoL scale. Only Physical Functioning and Role Functioning showed higher effect sizes in some comparisons. Responsiveness analyses were limited to one trial in advanced ovarian cancer showing sufficient change over time. In this subset, the Global Health/QoL scale was slightly more responsive than the Summary Score (d=−0.86 versus d=−0.81), while Appetite Loss showed the highest responsiveness (d=0.92).Conclusion: The Summary Score demonstrates strong construct validity and appears suitable as an overall HRQoL measure for distinguishing between clinically distinct groups. Evidence on responsiveness to change was limited in the present study and warrants further evaluation in settings where larger HRQoL changes are expected.

The European Organisation for Research and Treatment of Cancer (EORTC) QLQ-C30 is the most widely used patient-reported outcome measure in cancer clinical trials. Its multidimensional structure, comprising 15 scales and single items, requires careful selection of the most appropriate QLQ-C30 measure for a given study. Traditionally, the two-item Global Health Status/Quality of Life (Global Health/QoL) scale has served as the principal overall QLQ-C30 measure. The Summary Score was subsequently developed as an alternative aggregated measure reflecting symptoms and functioning across 13 QLQ-C30 scales. Because no single QLQ-C30 scale is optimal for all research settings, selection should be guided by the anticipated pattern of treatment effects. Individual QLQ-C30 scales remain the preferred measures when domain-specific treatment effects are hypothesized, whereas the Summary Score may be most appropriate when broad multidomain effects on HRQoL are expected. This paper presents guidance from the EORTC Quality of Life Group on when to use the Summary Score or the Global Health/QoL scale in different clinical research settings. The recommendations are based on a systematic review and comparative analysis of the literature and pooled analyses of randomized EORTC clinical trial data. The strongest evidence supports use of the Summary Score for cross-sectional analyses, where it consistently demonstrated greater discriminatory ability than the Global Health/QoL scale. Similarly, robust evidence supports its use for prognostic modelling of overall survival. In contrast, evidence regarding responsiveness to change over time is heterogeneous and does not support a clear preference for either the Summary Score or the Global Health/QoL scale.

Background: Hematologic toxicity is a major adverse event of temozolomide chemoradiotherapy for newly diagnosed glioblastoma. Although severe hematologic toxicity occurs in approximately 10% of patients, its impact on health-related quality of life (HRQoL) is unknown. This study evaluated the association between temozolomide-related hematologic toxicity and HRQoL in patients with newly diagnosed glioblastoma. Methods: Patients with newly diagnosed glioblastoma treated with temozolomide chemoradiotherapy and available HRQoL data from 3 clinical trials were included. Primary HRQoL outcomes were physical, social, and role functioning, pain, fatigue, nausea/vomiting, and global health status. Severe hematologic toxicity was defined as grade 3-4 neutropenia, lymphopenia, thrombocytopenia, or anemia; febrile neutropenia; or prolonged (>30 days) grade 2 toxicity. Associations between severe hematologic toxicity and HRQoL were evaluated using linear mixed models. Clinical relevance was interpreted as glioma-specific minimally important difference thresholds. Results: Among 1288 patients, 8% experienced severe hematologic toxicity during the concomitant phase, 6% during the maintenance and 1% during both. During the concomitant phase, severe hematologic toxicity was more frequent in females and in patients with MGMT promoter methylation. From baseline to week 6, global health mean scores declined by −8.52 points in patients with severe toxicity and −4.78 points in those without, with a non-significant difference (P = .24). Similar patterns were observed for physical, social, and role functioning. Symptom scales were numerically worse in severe toxicity but not significantly different. During the maintenance phase, differences were minor. Conclusion: Severe temozolomide-related hematologic toxicity showed no statistically significant association with impaired HRQoL, suggesting limited impact on patient-reported functioning and symptoms. Possible effects on psychological functioning should be examined further.

Background: This study presents the development of a glossary designed to harmonize terminology, foster consensus and support the successful implementation of the Setting International Standards in Analysing Patient-Reported Outcomes and Quality of Life Endpoints in Cancer Clinical Trials-Innovative Medicines Initiative (SISAQOL-IMI) guidelines. Methodology: SISAQOL-IMI consortium representatives from 41 international organisations were invited to participate in the development of the glossary. From 2021 to 2024, the glossary was developed through iterative processes, guided by consortium members input and formally updated twice yearly. A template with categories for term, acronym, topic, scientific definition, plain language definition, examples and references/sources was applied and were accessible for input throughout the process. As project documents such as draft guideline recommendations were updated, new terms for the glossary were identified. Definition of terms were primarily sought from a predefined hierarchical list of references and acknowledged resources, such as existing glossaries, prior to seeking other references. Patient representatives and patient advocates from The Workgroup of European Cancer Patient Advocacy Networks contributed to plain language definitions. Results: The glossary development resulted in 227 terms with scientific and plain language definitions, which are integrated into the online SISAQOL-IMI guidelines. During the initial development phase, of 205 terms, the project’s work package teams agreed on 166 (81%) and had diverging views on 39 (19%). Discussions were organized to reconcile disagreements. The glossary proved essential in 1) helping patient representatives better understand and engage in the consensus process, 2) aligning terminology during the pre-testing of preliminary recommendations, and 3) clarifying terms during the final language harmonisation of the recommendations. Conclusions: The glossary was pivotal in ensuring a common understanding of terminology throughout the SISAQOL-IMI project. It serves as a critical resource for enhancing clarity and successful implementation of the SISAQOL-IMI recommendations, ultimately advancing the role of patient-reported outcomes in improving the quality of cancer clinical trials.

PURPOSE: Prognosis in low-grade gliomas (LGGs) remains highly variable, and treatment-related late toxicity is a concern. This trial was comparing single-modality therapies in patients who often survive for years or decades and investigating differential responses according to molecular markers. METHODS: Four hundred seventy-eight patients with clinical high-risk LGG (WHO grade 2) were randomly assigned to standard radiotherapy (RT; 28 × 1.8 Gy) or dose-dense temozolomide (TMZ; 75 mg/m ² once daily × 21/28 days, up to 12 cycles). RESULTS: There was no significant difference in progression-free survival or overall survival (OS) between study arms. Analyzable tumor tissue in 73% (351/478) of patients allowed for post hoc reclassification according to the 2021 WHO pathologic criteria. In astrocytoma, IDHmt/1p/19q noncodeleted (n = 178), median OS was similar, 6.6-6.7 years irrespective of arm (0.67-1.44, P = .93). In oligodendroglioma, IDHmt/1p/19q codeleted (n = 109), median OS was 12.9 years (9.4-not reached) with RT and 14.9 years (10.1-number of events not reached) with TMZ (hazard ratio [HR], 0.88 [0.52-1.49], P = .63). In 64 tumors without isocitrate dehydrogenase (IDH) mutations, survival favored the TMZ arm: OS 2.5 (1.8-3.3) versus 4.7 (2.2-7.2) years (HR, 0.47 [0.27-0.82], P = .0068). Patients age 40 years and older fared better than patients younger than 40 years, challenging the current notion of age alone as a negative prognostic factor. CONCLUSION: The assigned initial treatment modality did not affect progression-free survival or OS, regardless of the molecular subtype. Combined-modality therapy was not tested in this trial but has since become a standard of care for IDH-mutant astrocytoma. With emerging novel therapeutic options, rational treatment strategies tailored at the individual clinical and pathologic recurrence risk profiles will be needed. The validity of an age cutoff as prognostic factor is challenged when tumors are molecularly classified.

BackgroundPatients with glioblastoma experience a substantial symptom burden that negatively affects functioning in daily life. Symptoms frequently co-occur and may form symptom clusters, yet the robustness, generalizability, and temporal stability of these clusters remain insufficiently studied. This is particularly relevant given the dynamic disease course and treatment-related effects in glioblastoma.MethodsWe used data from the CODAGLIO 2.0 database, comprising adult patients with newly diagnosed or recurrent glioblastoma from seven randomized controlled trials not included in previous clustering studies. Symptoms were assessed using 18 symptom scales from EORTC QLQ-C30 and QLQ-BN20 questionnaires at baseline and 3-month follow-up. Symptom prevalence and multi-symptom co-occurrence were described. Symptom clusters were identified using Spearman correlations, partial correlations, and hierarchical cluster analysis and compared with previously reported clusters. Cluster stability over time was evaluated using cophenetic correlation, tanglegrams, entanglement, and adjusted Rand index scores.ResultsAmong 1,795 patients with baseline HRQoL data, 1678 patients (94%) reported multi-symptom co-occurrence at baseline and 1045 patients (96%) at follow-up. Fatigue, communication deficits, motor dysfunction, drowsiness, insomnia, and visual disorder were most prevalent. Four clinically meaningful symptom clusters were identified: pain–headache, motor dysfunction–weakness of the legs, fatigue–drowsiness, and a seizure-related cluster. These clusters largely aligned with previously reported findings and showed moderate stability over time (cophenetic correlation 0.84; adjusted Rand index 0.41), with greatest variability observed in seizure-related symptoms.ConclusionSymptom clusters in glioblastoma are robust yet dynamic. Incorporating cluster-based and longitudinal symptom assessment may enhance symptom monitoring and support more personalized supportive care strategies.

Supplementary Figure 5: Copy number variants in the three main glioblastoma groups according to classifier v12.5. A Copy number variants in all patients combined (n = 295), B in RTK1 tumors (n = 89), C in RTK2 tumors (n = 140) and D in MES tumors (n = 66). PDGFRA amplifications are enriched in RTK1 tumors and gains of chromosomes 19 and 20 are enriched in RTK2 tumors.

Supplementary Table 1: Representativeness of study participants.

Supplementary Figure 1: Overview of the clinical trial and biomarker analysis. Out of 596 patients in the EORTC-26101 trial, patient tumor tissue to perform DNA methylation (Meth) and NGS sequencing analysis was available from 380 patients (63.8%). Of these, 99 patients treated with lomustine and 189 patients treated with lomustine and bevacizumab were eligible to detect biomarkers for bevacizumab response.