Long-term neurocognitive function and quality of life after multimodal therapy in adult glioma patients: a prospective long-term follow-up

Abstract

Purpose: Multimodal therapies have significantly improved prognosis in glioma. However, in particular radiotherapy may induce long-term neurotoxicity compromising patients' neurocognition and quality of life. The present prospective multicenter study aimed to evaluate associations of multimodal treatment with neurocognition with a particular focus on hippocampal irradiation.

Methods: Seventy-one glioma patients (WHO grade 1-4) were serially evaluated with neurocognitive testing and quality of life questionnaires. Prior to (baseline) and following further treatment (median 7.1 years [range 4.6-11.0] after baseline) a standardized computerized neurocognitive test battery (NeuroCog FX) was applied to gauge psychomotor speed and inhibition, verbal short-term memory, working memory, verbal and non-verbal memory as well as verbal fluency. Mean ipsilateral hippocampal radiation dose was determined in a subgroup of 27 patients who received radiotherapy according to radiotherapy plans to evaluate its association with neurocognition.

Results: Between baseline and follow-up mean performance in none of the cognitive domains significantly declined in any treatment modality (radiotherapy, chemotherapy, combined radio-chemotherapy, watchful-waiting), except for selective attention in patients receiving chemotherapy alone. Apart from one subtest (inhibition), mean ipsilateral hippocampal radiation dose > 50 Gy (Dmean) as compared to < 10 Gy showed no associations with long-term cognitive functioning. However, patients with Dmean < 10 Gy showed stable or improved performance in all cognitive domains, while patients with > 50 Gy numerically deteriorated in 4/8 domains.

Conclusions: Multimodal glioma therapy seems to affect neurocognition less than generally assumed. Even patients with unilateral hippocampal irradiation with > 50 Gy showed no profound cognitive decline in this series.

Keywords: Glioma; Multimodal tumor-directed treatment; Neurocognition; Quality of life; Radiotherapy.

Fig. 1

Flow chart of patients included in the present series

Fig. 2

Mean cognitive performance (in z-scores) and standard deviations (error bars) in NeuroCog FX subtests at baseline and follow-up (median 7.1 years [range 4.6–11.0] after baseline), separated for treatment groups. a Watchful-waiting (n = 24), b Chemotherapy (n = 11), c Radiotherapy (n = 7), d Combined radio-chemotherapy (n = 29). Statistically significant changes in cognitive performance between baseline and follow-up are indicated by asterisks (*p < .05, **p < .01, ***p < .001)

Fig. 3

Cognitive performance in relation to hippocampal dosimetry. a Change in cognitive performance according to mean radiation dose on ipsilateral hippocampus on individual patient level. Z-scores > 0 indicate an improvement of cognitive performance, z-scores < 0 indicate a deterioration of cognitive performance. Individual data only refer to digital RT plans with exact hippocampal dosage in Gy available for patient groups with Dmean at ipsilateral Hippocampus < 10 Gy vs. > 50 Gy b Change of cognitive functioning in the long-term on group level. Graphs indicate mean cognitive performance (z-scores) with standard deviations represented by error bars in NeuroCog FX subtests at baseline (i.e., after surgery) and at follow-up (median 7.1 years [range 4.6–11.0] after baseline), separated for patients with Dmean at ipsilateral Hippocampus < 10 Gy (n = 8) and patients with Dmean at ipsilateral Hippocampus > 50 Gy (n = 12). The asterisk indicates a statistically significant interaction effect of timepoint and group (*p < .05)