Longitudinal Assessment of Neurocognitive Outcomes and Correlation with Limbic System Radiation Doses in Patients Undergoing Radiotherapy for Central Nervous System Tumours: A preliminary report

Introduction

The prevalence of cranial nervous system tumours has increased due to improved diagnosis, treatment protocols, and multimodal treatment approaches. According to GLOBOCAN 2022 data, the global incidence of Central Nervous System (CNS) tumours is approximately 322,000.¹ As per National Cancer Registry Program (NCRP), the Indian data records the incidence of CNS tumours ranging from 5 to 10 cases per 100,000 population, accounting for approximately 2% of all malignancies.²

The management of gliomas presents a unique challenge, not only due to their aggressive biological behaviour but also because of their profound effects on neurocognitive function. As advancements in neurosurgical techniques, chemotherapy, and radiotherapy have improved survival rates, the focus has increasingly shifted toward preserving the quality of life, particularly cognitive function. Neurocognitive impairment is a common finding in glioma patients, impacting functions such as memory, executive function, attention, and language function^3,4 While surgical resection, chemotherapy, and radiotherapy are the mainstay modalities of treatment, the effect of these modalities on neurocognitive function is a significant area of research. Specifically, radiation therapy has been associated with cognitive impairment, most probably because of radiation-induced damage to critical cognitive and limbic system structures of the brain such as the hippocampus, corpus callosum, and amygdala^5,6 New radiation techniques such as volumetric-modulated arc therapy (VMAT) have been developed to better localize the tumour with less exposure to normal neural tissue, potentially reducing neurocognitive toxic effects.⁷

The neurocognitive effects of radiation are multifactorial, involving direct damage to neurons, vascular damage, and neuroinflammation. The extent of cognitive impairment depends on factors like tumour location, radiation dose, fractionation schedules, and patient-specific vulnerability.^6,8 Previous studies have highlighted the importance of assessing cognitive outcomes following radiotherapy using standardized neurocognitive tests.^9,10 The Addenbrooke’s Cognitive Assessment (ACA) Scale is one such validated instrument for assessing various cognitive domains, including memory, attention, fluency, language, and visuospatial processing, and is a valuable instrument for glioma patients undergoing radiotherapy.¹¹

Several studies have previously considered neurocognitive outcomes after surgery and radiotherapy for glioma patients.^6,12 Talacchi et al. described that tumour resection by itself can be associated with cognitive change, irrespective of adjuvant treatments. Although the influence of the dose of radiation on the remaining limbic structures and the cerebellum is underrepresented in the literature, hippocampal dose constraint studies have proved that lowering the dose of radiation to the hippocampus can spare cognition, reaffirming the position of accurate radiation planning.⁵

The purpose of this study was to examine the correlation of dose delivered to the critical limbic system structures—hypothalamus, corpus callosum, amygdala—and anterior and posterior cerebellum, and resulting neurocognitive change measured by the ACE Scale.

Methodology

This prospective observational study was conducted to assess early neurocognitive changes in patients with primary central nervous system (CNS) tumours receiving radiation therapy and to evaluate the correlation between cognitive outcomes and radiation dose to specific brain substructures involved in cognition. The study received approval from the institutional ethics committee, and written informed consent was obtained from all participants prior to start of the study.

Adult patients (≥18 years) with newly diagnosed, histologically confirmed gliomas of WHO Grade II to IV were included. Eligibility criteria required a Karnofsky Performance Score (KPS) of seventy or above. Patients were excluded if they had prior cranial radiotherapy, pre-existing cognitive impairment, neurodegenerative disorders, psychiatric illness affecting cognition, history of traumatic brain injury, or other neurological conditions known to affect cognitive function. Recurrent or residual tumours from prior treatments were also excluded.

Demographic and clinical data were collected at baseline, including age, sex, socioeconomic indicators. Tumour-related data included anatomical location, laterality, WHO grade, histological subtype, and molecular markers. Extent of surgical resection was categorized as biopsy, subtotal resection, or gross total resection.

Results

Discussion

Neurocognitive function has become an essential parameter in the management of primary CNS malignancies, especially as survival improves with multimodal treatments. Cognitive impairment in these patients arises not only from the tumour itself but also from interventions such as surgery, chemotherapy, and radiation therapy. This prospective study aimed to evaluate neurocognitive decline following focal radiotherapy (RT) in glioma patients using the Addenbrooke’s Cognitive Examination III (ACE-III), and to explore associations between cognitive change and radiation dose to key neuroanatomical structures. A statistically significant decline in total ACE-III scores was observed post-radiotherapy, with mean scores decreasing from 89.48 ± 6.84 to 87.91 ± 6.29 (p = 0.007). Domain-wise, attention, language, and fluency were the most affected, while memory and visuospatial abilities were preserved. Notably, patients who exhibited greater cognitive decline tended to have received higher doses to key neural structures, particularly the corpus callosum.

The ACE-III tool enabled domain-specific assessment of cognitive change, capturing subtle yet meaningful declines in attention, language, and fluency. Compared to tools like the MMSE (Mini-Mental State Examination) and HVLT-R (Hopkins Verbal Learning Test-Revised), the ACE-III offers broader coverage and greater sensitivity to early changes.¹¹ This allowed for more precise mapping of cognitive changes to anatomical structures.

These findings align with the growing understanding that radiotherapy to limbic system structures can lead to early cognitive impairment when critical neural structures are exposed to moderate-to-high doses.¹⁶ While hippocampal avoidance has been the focus of most cognitive-sparing radiotherapy studies, including the NRG-CC001 trial which demonstrated memory preservation with hippocampal-sparing WBRT plus memantine, our data highlight that attention and fluency were more prominently affected domains.¹⁷ This discrepancy may reflect the focal, rather than whole-brain, nature of RT in glioma patients.

The corpus callosum, a major interhemispheric white matter tract responsible for integrating cortical functions and sustaining attention, showed a strong trend toward correlation with cognitive decline in our cohort. Although statistical significance was not reached across all metrics, higher doses (>35 Gy) to the corpus callosum were associated with greater reductions in ACE-III scores. This is supported by the study by Redmond et al., who demonstrated that radiation exposure to white matter tracts such as the corpus callosum and superior longitudinal fasciculus predicted neurocognitive dysfunction in glioma patients.¹⁸ Our findings are further corroborated by Chapman et al., who used diffusion tensor imaging to show microstructural degeneration of the corpus callosum associated with attention and processing speed deficits.¹⁹

Of the 25 patients, 7 (28%) had abnormal ACE-III scores at baseline post-surgery. In contrast, 8 (32%) had abnormal scores at 6 months post-radiotherapy. This corresponds to a 4% increase in cognitive dysfunction after radiotherapy. The 28% baseline impairment reflects tumour- and surgery-related cognitive effects. Hence it may not be justified to refer radiotherapy as the primary cause for the observed cognitive decline, and therefore the results should be interpreted keeping in mind other possible clinical factors and outcomes.

Although the hippocampus remains central to memory preservation strategies, our study did not find a statistically significant relationship between hippocampal dose and post-treatment memory decline. This contrasts with results from Gondi et al., who reported hippocampal dose-dependent memory loss, particularly in low-grade or benign intracranial tumours.²⁰ The absence of a similar association in our study may be attributed to lower hippocampal doses overall, contralateral compensation in unilateral lesions, or sample size constraints. Our study also examined radiation exposure to other critical brain structures. The amygdala and cerebellum, although not primary targets, received variable radiation doses depending on tumour location. Although no significant correlation with cognitive outcomes was noted, prior studies suggest these structures contribute to cognitive-affective integration and executive function.²¹ A few Studies in literature have similarly identified radiation dose to frontal and temporal lobes as predictors of decline in language and fluency.^22,23 The findings also resonate with Haisraely et al., who proposed expanding dose-sparing protocols beyond the hippocampus to include white matter tracts.²⁴ Our results support this direction, showing that cognitive impairment involves a distributed neural network rather than isolated memory pathways alone. Talacchi et al. reported that cognitive diminution post-surgery can occur in up to 30–40% of patients, and more likely in executive functions, memory, and attention, domains largely mediated by frontal-subcortical networks.⁴ As our study population had higher number of frontal lobe lesions, the cognitive impairment that we observed may reflect the impact of surgery on cognition.

Despite the consistency with literature, our study has few limitations. The current analysis includes a smaller cohort and a short follow up time, which may have obscured significant dose-response associations. However, patient recruitment is ongoing, and future analyses with a larger sample size and extended follow-up are expected to strengthen the findings. Functional neuroimaging techniques, such as volumetric MRI or diffusion tensor imaging, were not incorporated at this stage, which may have further elucidated the structural basis of observed cognitive changes. However, patient recruitment is ongoing, and future analyses with a larger sample size and extended follow-up are expected to strengthen the findings and allow for more robust dose-volume-outcome correlations.

Future studies should include larger, multicentric cohorts with longitudinal cognitive assessment and advanced neuroimaging. Expanding cognitive-sparing strategies to include midline white matter tracts, cerebellar structures, and language circuits may improve outcomes.

Conclusion

This study demonstrates that early neurocognitive decline following focal RT exists, particularly affecting attention and language. While hippocampal dose remains relevant, the corpus callosum appears to play a critical role in early cognitive impairment. These findings underscore the need for comprehensive neuroanatomical preservation in radiotherapy planning.

Consent

Written informed consent to take part in the research has been obtained from all patients when indicated. Consent for publication is not applicable as all data is anonymized, and no images are being published.

Ethics statement

The Institutional Ethics Committee of Kasturba Medical College approved this study, Mangalore, vide protocol no. IEC KMC MLR 09/2024/554 dated 19.09.2024. The study was conducted in accordance with the ethical guidelines and regulations set forth by the institution. The study has been conducted according to the principles expressed in the Declaration of Helsinki.

Author contributions

All authors contributed equally to the conceptualization, methodology, data collection, analysis, manuscript writing, and revision. All authors have read and approved the final version of the manuscript.