From Metabolism to Mood: Imaging the Neuropsychiatric Consequences of Chemotherapy-Induced Brain Changes

Inclusion criteria

• Patients with histopathologically proven cancer who had been referred to the

Department of Nuclear Medicine for baseline and response evaluation PET/CT

• Age ≥ 18 years

• Patients who had been planned for a baseline PET/CT, 2 cycles of systemic

chemotherapy followed by a response evaluation PET/CT

• Patients who could read and write either Hindi or English

• Had provided informed consent

Exclusion criteria

• History of prior chemotherapy

• Prior RT to the head and neck region

• Previously diagnosed with neurological/psychiatric disorder or intracranial lesions.

Neuropsychiatric Assessment was conducted at two time points- before both baseline and response evaluation ¹⁸F FDG PET/CTs were conducted. The assessment was made by a clinical psychologist and a consultant psychiatrist using structured clinical interview as well as validated questionnaires – Montral Cognitive Assessment – Basic (MoCA B), 14 item Hamilton Anxiety Rating scale (HAM A )and 17 item Hamilton Depression Rating scale (HAM D) [Supplementary Appendix 1-3].

Imaging and interpretation

Patients were required to fast for 6 hours before the radiopharmaceutical injection. On arrival, patients were injected with 0.14- 0.2 mci/kg of ¹⁸F FDG, provided their blood glucose was below 200 mg/dl. After this, patients were seated in a quiet, dimly lit environment with closed eyes. Forty-five minutes post-injection, the patient was taken up for a scan. PET/CT was performed by institutional protocol using an integrated Philips Ingenuity 128 slice TOF PET/CT scanner. A dedicated 5-minute brain spot was acquired before the routine whole-body imaging. Following this, images were fused using the Philips workstation and interpreted by an experienced nuclear medicine physician. 3-6 weeks after the second cycle of chemotherapy, patients underwent a response evaluation with a PET/CT scan with a similar protocol. Metabolism in different brain regions was assessed visually, and the assessment was further complemented by NeuroQ software. Brain ¹⁸F FDG PET findings were used solely for research and correlative analysis and did not influence oncological treatment decisions.

NeuroQ software analysis

It is used to assess neurodegenerative processes and underlying symptoms of cognitive and movement disorders by comparing regional activity values to each other and to those in brain scans acquired for asymptomatic control subjects. In our study, the brain PET scans were normalised to the Whole Brain. The NeuroQ software conducts automatic segmentation into the brain regions as listed in Supplementary Table 1 and demonstrated in Supplementary Fig 1. The Z score of each lobe of the brain was estimated by calculating the average of each of the subregions.

Data analysis

MS Excel was used for data entry. Statistical R software was used for data analysis. Quantitative variables such as Z scores of different brain regions before and after chemotherapy were described using the mean, standard deviation, maximum, and minimum values. For ordinal data such as severity of depression, anxiety and cognitive impairment, frequency and percentage were calculated. The Shapiro-Wilk test was used for normality check of the pre- and post-chemotherapy neuropsychiatric scores. The Wilcoxon signed rank test was used to analyse the difference between pre- and post-chemotherapy neuropsychiatric scores. Man- Whitney U test was used to analyse the correation between sex, metastatic status and neuropsychiatric scores. Spearman correlation was used to analyse the correlation between age and neuropsychiatric scores and between the difference in Z scores and neuropsychiatric scores pre- and post-chemotherapy. Multiple linear regression analysis was performed to assess the effect of variables such as age, sex, metastatic status and Z score changes on neuropsychiatric outcomes. A p-value < 0.05 was considered statistically significant.

A schematic representation of the research methodology is presented in Supplementary Fig 2.

Severity of anxiety and depression

Before chemotherapy, 19.6% of patients had no anxiety, but this dropped to 1.7% post-chemotherapy. Similarly, the proportion of patients with mild anxiety decreased from 48.21% to 32.14%, while those experiencing mild to moderate anxiety increased from 23.21% to 33.92%. Most notably, moderate to severe anxiety cases rose from 8.9% to 32.14%, indicating a significant shift towards higher anxiety levels after chemotherapy. These findings suggest a worsening of anxiety symptoms post-treatment, with more patients progressing to moderate or severe anxiety. The Wilcoxon signed rank test confirmed that the shift in anxiety severity before and after chemotherapy was statistically significant (p < 0.05) [Table 2, Supplementary Fig 4].

Before chemotherapy, 26.78% of patients had no depression, but this sharply declined to 3.57% post-chemotherapy. Similarly, the proportion of patients with mild depression dropped from 16.07% to 1.7%, while moderate depression cases remained relatively stable (16.07% pre vs. 10.71% post). However, a significant shift was observed in very severe depression cases, which increased dramatically from 21.42% to 75% post-chemotherapy. The Wilcoxon signed rank test confirmed that the shift in depression severity before and after chemotherapy was statistically significant (p < 0.05) [Table 3, Supplementary Fig 5].

Severity of cognitive abnormalities

Before chemotherapy, 50% of patients had normal cognition (MoCA ≥26), but none remained in this category post-chemotherapy. Instead, the proportion of mild cognitive impairment cases increased from 50% to 64.28%, while moderate cognitive impairment emerged in 26.78% of patients. No cases of severe cognitive impairment were observed. This suggests a clear deterioration in cognitive function after chemotherapy, with all patients shifting from normal cognition to at least mild impairment and a notable proportion developing moderate cognitive impairment. The Wilcoxon signed rank revealed a statistically significant shift from normal to impaired cognition (p < 0.05) [Table 4, Supplementary Fig 6].

Correlation between demographic variables and change in neuropsychiatric scores

No significant correlation was noted between the age, sex and metastases status of the patients with the changes in neuropsychiatric scores [Supplementary table 2].

Correlation between change in Z-scores in different regions of the brain and neuropsychiatric scores:

Frontal Z-score changes showed a strong positive correlation with MoCA scores (ρ = 0.77, p < 0.05), indicating that cognitive decline is linked to reduced frontal metabolism. Temporal Z-score changes were positively correlated with HAM-A (ρ = 0.6, p < 0.05) and HAM-D (ρ = 0.75, p < 0.05), suggesting that anxiety and depression worsen with decreased temporal metabolism. Other brain regions, including the parietal lobe, basal ganglia, brainstem, cerebellum, and occipital lobe, did not show statistically significant correlations (p > 0.05). These findings highlight the neurobiological basis of chemotherapy-induced cognitive and emotional dysfunction, particularly involving the frontal and temporal lobes [Supplementary Table 3, Supplementary Fig 7 and 8].

Impact of brain glucose metabolism on neuropsychiatric outcomes

Multiple linear regression analysis was performed to assess the relationship between chemotherapy-induced brain metabolism changes and neuropsychiatric outcomes, including anxiety (HAM-A), depression (HAM-D), and cognitive function (MoCA). The model included age, sex, metastatic status, and Z-score changes in multiple brain regions as independent variables. Fig 1 and 2 illustrate the concordance between regional alterations in cerebral glucose metabolism and the degree of deterioration of neuropsychiatric status.

Close

Fig 1:

A 47 year old female patient with metastatic breast cancer, (A) Pre-chemotherapy brain PET images, (B) Pre- chemotherapy brain PET NeuroQ Analysis, (C) Post-chemotherapy brain PET images, (D) Post- chemotherapy brain PET NeuroQ Analysis. Post chemotherapy images show significant hypometabolism in temporal lobes as compared to the pre- chemotherapy images- as noted in the sections highlighted by the red boxes in (A) and (C) and yellow block arrows in (B) and (D). Neuropsychiatric evaluation revealed a significant deterioration of neuropsychiatric status, particularly depression and anxiety scores with pre-chemotherapy scores of MoCA B-27, HAM A- 14 and HAM D- 8 and post-chemotherapy scores of MoCA B- 26, HAM A- 28 and HAM D- 18. Patient had a partial metabolic response to chemotherapy. Colour coded Z-score maps were used by NeuroQ software to depict the degree of hypometabolism with warmer colour (pink/red) indicating more severe hypometabolism and cooler colour (blue) indicating milder hypometabolism. (Numbers in the images refer to the slice numbers as provided by the NeuroQ software). PET: Positron emission tomography; MoCA: Montreal cognitive assessment; HAM-A: Hamilton anxiety scale; HAM-D: Hamilton depression scale

Export to PPT

Close

Fig 2:

A 68 year old female patient with non-metastatic locally advanced breast cancer (A): pre-chemotherapy brain PET images, (B) Prechemotherapy brain PET NeuroQ Analysis and (C) Post-chemotherapy brain PET images, (D) Post- chemotherapy brain PET NeuroQ Analysis. Post chemotherapy images show mild hypometabolism in right frontal lobes- as indicated by sections highlighted in the red boxes in (A) and (C) and yellow block arrows in (B) and (D). Neuropsychiatric evaluation revealed only a mild deterioration of neuropsychiatric status with MoCA B - 27, HAM A – 16 and HAM D – 8 and post-chemotherapy scores of MoCA B - 23 , HAM A - 14 and HAM D- 10. Colour coded Z-score maps were used by NeuroQ software to depict the degree of hypometabolism with warmer colours (pink/red) indicating more severe hypometabolism and cooler colours (blue) indicating milder hypometabolism. (Numbers in the images refer to the slice numbers as provided by the NeuroQ software). PET: Positron emission tomography; MoCA: Montreal cognitive assessment; HAM-A: Hamilton anxiety

Export to PPT

• Cognitive Decline and Frontal Lobe Metabolism: A significant association was found between a greater difference in frontal lobe metabolism and a decline in MoCA scores (β = -0.7, p = 0.005), suggesting that chemotherapy-induced metabolic reductions in the frontal lobe contribute to cognitive impairment.

• Anxiety and Temporal Lobe Metabolism: A greater difference in temporal lobe metabolism was significantly correlated with increased HAM-A scores (β = 0.6, p = 0.02), indicating that chemotherapy-related reductions in temporal lobe activity may exacerbate anxiety symptoms.

• Depression and Temporal Lobe Metabolism: Similarly, a greater difference in temporal lobe metabolism was significantly correlated with increased HAM-D scores (β = 0.8, p = 0.01), suggesting that chemotherapy-related metabolic changes in this region contribute to depressive symptoms [Supplementary Fig 9].

Hamilton anxiety scale (HAM A) analysis

The Hamilton Anxiety Scale (HAM-A) analysis revealed a marked increase in the proportion of patients with moderate to severe anxiety, from 8.9% before chemotherapy to 32.14% after treatment. This rise underscores the psychological burden of chemotherapy. Similar results were observed by Horky et al. , who reported that up to 50% of patients experienced anxiety following chemotherapy^[8], and by Dietrich et al., particularly among those receiving platinum-based regimens known for neurotoxic effects.^[9] Schroyen et al. also documented that 23% of breast cancer patients undergoing chemotherapy exhibited anxiety symptoms, significantly higher than in the general population.^[10]

The increase in anxiety can be attributed to both neurophysiological effects and the emotional stress of cancer treatment, including fear of disease progression and treatment-related complications. These findings highlight the importance of routine psychological assessment and interventions such as cognitive-behavioural therapy to reduce chemotherapy-induced anxiety and improve mental health outcomes.

Hamilton depression scale (HAM-D) analysis

A substantial worsening of depressive symptoms was also observed after chemotherapy. The percentage of patients with no depression decreased from 26.78% pre-chemotherapy to 3.57% post-chemotherapy, while those with very severe depression rose to 75%. These findings are consistent with previous studies. Gholami et al. reported increased depression severity following chemotherapy, especially in patients on intensive regimens ^[11], while Dietrich et al. described depression as one of the most common psychological sequelae of chemotherapy.^[9] The mechanisms underlying these changes likely include both neurochemical alterations in the brain and the psychological distress associated with treatment.

Machado et al. emphasised that untreated depression adversely affects treatment adherence and quality of life in cancer patients.^[12] The present findings reinforce the necessity for integrating mental health care within oncology services to identify and manage depression early in the treatment course.

Montreal cognitive assessment (MoCA) analysis

The Montreal Cognitive Assessment (MoCA) results demonstrated a significant decline in cognitive function post-chemotherapy. The proportion of patients with normal cognitive function (scores ≥26) dropped from 50% before treatment to 0% afterward, while moderate cognitive impairment increased to 26.78%. This pattern aligns with the well-documented phenomenon of chemotherapy-induced cognitive dysfunction, commonly termed “chemo brain.”

Horky et al. reported reduced glucose metabolism in the frontal cortex and basal ganglia, regions essential for executive processing and working memory.^[8] Similarly, Hu et al. demonstrated a decline in glucose metabolism in the frontal lobes and hippocampus, structures vital for cognitive performance.^[13] These studies, in conjunction with the present findings, suggest that chemotherapy-induced metabolic disruptions contribute to impairments in memory, attention, and executive functions, substantially affecting post-treatment quality of life. Cognitive rehabilitation and neuroprotective strategies should therefore be considered for patients at risk.

Z-score analysis of brain regions

Descriptive analysis of Z-scores revealed significant post-chemotherapy reductions in glucose metabolism in the frontal and temporal regions. These results are consistent with prior studies showing chemotherapy-induced alterations in brain metabolism. Gamal et al. reported a 20–25% decrease in FDG uptake in the frontal and mesial temporal lobes ^[14], while Schroyen et al. observed similar reductions in these regions.^[10] The frontal and temporal lobes are integral to executive function, emotional regulation, and memory; thus, metabolic impairment in these areas likely underpins the observed cognitive and psychiatric symptoms.

The selective vulnerability of these regions may stem from their high metabolic demand and dense synaptic activity, making them more susceptible to chemotherapeutic neurotoxicity. Future investigations should explore the mechanisms behind these alterations and evaluate potential neuroprotective interventions or cognitive training methods to preserve brain function during therapy.

Correlation between Z-scores and neuropsychiatric parameters

Correlation analysis demonstrated significant positive relationships between changes in Z-scores of the frontal and temporal regions and alterations in cognitive function (MoCA), anxiety (HAM-A), and depression (HAM-D). Notably, changes in temporal region Z-scores showed strong correlations with both anxiety (ρ = 0.6, p < 0.05) and depression (ρ = 0.75, p < 0.05). These findings indicate that chemotherapy-induced metabolic reductions in key brain regions are closely associated with neuropsychiatric deterioration.

Shrot et al. found similar associations between metabolic changes and psychological symptoms in paediatric cancer patients undergoing chemotherapy [15], while Gamal et al. linked alterations in brain glucose metabolism with neuropsychological outcomes^[14]. The temporal lobe, responsible for mood regulation and memory, appears particularly susceptible to neurotoxic effects, explaining its strong correlation with increased anxiety and depression scores.

These results highlight the interconnection between cerebral metabolism, cognition, and mental health. The observed correlations suggest that addressing both cognitive and affective symptoms simultaneously could improve overall well being in cancer patients receiving chemotherapy. Incorporating neuroimaging biomarkers into routine follow-up may aid in early detection of neuropsychiatric side effects and guide the development of targeted interventions.

Routine neuropsychiatric screening should be integrated into oncology follow ups to enable early detection and management of anxiety, depression, and cognitive deficits. Selective use of ¹⁸F FDG PET/CT imaging in patients showing early cognitive changes may aid personalised care and neuroprotection. A multidisciplinary framework combining medical, psychological, and rehabilitative interventions is recommended. Future trials of chemotherapeutic agents should include neurocognitive and imaging endpoints to define neurotoxicity profiles and explore protective strategies such as drugs, exercise or mindfulness therapies. Exploration of genetic predispositions and neuroprotective interventions may further guide individualised management strategies. Further, the effect of number and duration of chemotherapy cycles, existing comorbidities and tolerability to cytotoxic regimes should also be studied in further research.

As per our knowledge, there are no prior prospective studies that have included both neuropsychiatric clinical evaluation and metabolic imaging. This study’s primary strength lies in its integrative design, combining ¹⁸F FDGPET imaging with validated neuropsychiatric assessments to evaluate the multidimensional effects of chemotherapy on brain metabolism, cognition and mental health. The longitudinal pre- and post-chemotherapy assessment allows direct comparison of metabolic and psychological changes, strengthening the evidence for chemotherapy-induced neurotoxicity.