Transcranial direct current stimulation enhances cognitive performance and modulates QEEG beta- and theta-band power in patients undergoing whole-brain radiotherapy Authors: Leila Halavi Anjar, Amir Ghasemi Jangjoo, Touraj Hashemi, Davood Khezerloo, Hassan Sabouri Moghaddam Published: 2026 ABSTRACT AbstractObjectiveTo investigate whether transcranial direct current stimulation (tDCS) can mitigate radiotherapy-induced cognitive impairment and modulate cortical activity, as measured by quantitative electroencephalography (QEEG), in patients undergoing whole-brain radiotherapy (WBRT).MethodsIn this randomized, sham-controlled pretest–posttest study, 40 right-handed adults with brain metastases scheduled for WBRT were assigned to either active tDCS (n = 20) or sham stimulation (n = 20). For tDCS stimulation (2 mA, 20 sessions), two electrode montages were applied over the dorsolateral prefrontal cortex (DLPFC) and supraorbital regions. Cognitive performance was assessed using a computerized 2-back task (working memory) and the continuous performance test (CPT) for sustained attention. Resting-state QEEG was recorded before and after treatment to quantify absolute beta and theta power. Peripheral blood samples were analyzed by Western blot to evaluate expression of DYRK1A, MMP2, MMP9, and VEGF. Multivariate ANCOVA was used to control for baseline differences.ResultsActive tDCS significantly improved cognitive performance compared with sham stimulation. N-back correct response increased from 2.96 ± 1.53–4.75 ± 0.55 (about 60.47% improvement, p < 0.001), and CPT correct responses increased from 13.20 ± 2.52–18.80 ± 1.39 (about 42.42% improvement, p < 0.01). QEEG analysis showed a significant reduction in absolute beta power (17.41 ± 4.30–11.46 ± 3.15 μV², about 51.19% reduction, p < 0.001), and theta power (33.41 ± 5.20–18.15 ± 4.50 μV², about 84.07% reduction, p = 0.002). Western blot results indicated that tDCS significantly attenuated the post-radiotherapy upregulation of stress- and inflammation-related proteins.ConclusiontDCS was associated with improved short-term cognitive performance during WBRT and may improve working memory and sustained attention, normalize aberrant beta–theta EEG activity, and reduce the expression of peripheral biomarkers. These findings support tDCS as a potential noninvasive adjunctive strategy for preserving cognitive function in patients undergoing WBRT.Keywords: Transcranial direct current stimulation, Brain metastases, Radiotherapy, Cognitive rehabilitation, Quantitative electroencephalography Introduction Whole-brain radiotherapy (WBRT) is one of the main treatment options for certain primary and metastatic brain tumors, as well as a prophylactic therapy for several cancers such as acute lymphoblastic leukemia and small cell lung cancer (Protopapa et al., 2019, Mantovani et al., 2021). Although this treatment often prolongs survival, it frequently does so at the expense of cognitive health. Patients may experience acute deficits in verbal learning and working memory, early delayed declines in attention, and late-delayed, irreversible impairments associated with microvascular injury, demyelination, stem cell loss, and direct neuronal damage (Viner, 2009, Yang et al., 2015). Quantitative electroencephalography (QEEG) can be used as a noninvasive method to evaluate radiotherapy-induced brain damage (Kropotov, 2010, Duff, 2004). By analyzing the absolute beta- and theta-band power during the resting state, it is possible to detect signatures of disrupted neural synchrony that correlate with impairments in executive function and working memory (Wianda, 2019, Uhlhaas and Singer, 2006). Transcranial direct current stimulation (tDCS) is a safe and well-tolerated neuromodulation technique capable of modulating cortical excitability and enhancing neuroplasticity (Saran, 2024, Rozisky et al., 2016). In this technique, a weak electrical current is delivered through two electrodes (target and reference) placed on the patient’s scalp (Khadka et al., 2019, Thair et al., 2017). tDCS can be applied to the dorsolateral prefrontal cortex (DLPFC)—an area central to attention, memory, and executive control—where it enhances synaptic efficacy and dendritic connectivity without major side effects (Li et al., 2022). tDCS has been investigated as a noninvasive neuromodulatory intervention across a range of neurological and psychiatric conditions, although its regulatory status and approved indications vary by device and jurisdiction. This method has regulatory clearance for specific clinical indications such as major depressive disorder, while cognitive applications remain investigational. Several studies have evaluated the therapeutic and rehabilitative effects of tDCS on various types of brain injury, including traumatic brain injury, pediatric brain injury, and stroke (Zaninotto et al., 2019, Cerezo-Zarzuelo et al., 2025, Trofimov et al., 2018, Bornheim et al., 2022). These studies have demonstrated improvements in Coma Recovery scales among patients with disorders of consciousness, as well as enhanced cognitive function (Bender et al., 2023, Cortese et al., 2015). Some investigations have also reported neurophysiological changes that correlate with clinical improvements (Zaninotto et al., 2019). Moreover, multiple studies have described beneficial effects of tDCS in managing pain among patients receiving radiotherapy for head and neck cancers (de Souza Moura et al., 2022, Hu et al., 2016). However, to date, no study has examined the effects of tDCS on cognitive performance and QEEG beta- and theta-band power in patients with brain metastases undergoing WBRT. Therefore, the current study aimed to evaluate whether repeated sessions of tDCS applied over the DLPFC and supraorbital regions could mitigate radiotherapy-related cognitive decline in patients with brain metastases undergoing WBRT. We assessed changes in attention and working memory, together with changes in absolute beta- and theta-band power measured by QEEG. Materials and methods Study design This study was designed as a randomized, sham-controlled pretest–posttest trial. Forty patients with brain metastases scheduled to undergo WBRT between 2024 and 2025 were randomly assigned in equal numbers to receive either active tDCS or sham (control) stimulation. Participants were randomly assigned (1:1) to the active tDCS or sham group using a computer-generated randomization sequence prepared by an investigator who was not involved in outcome assessments. Allocation assignments were placed in sequentially numbered, opaque, sealed envelopes and were opened by the study coordinator only after participant enrollment. Eligible participants were enrolled by the treating radiation oncologist after confirmation of the inclusion and exclusion criteria. Group allocation was performed by a study coordinator who opened the sealed envelopes after enrollment. Participants, investigators performing cognitive assessments, QEEG analysts, and laboratory personnel responsible for Western blot analyses remained blinded to treatment allocation throughout the study. The statistician remained blinded to group allocation and performed all analyses using coded datasets. Cognitive assessments and QEEG recordings were performed immediately before the first radiotherapy session (pretest) and immediately after completion of the final tDCS session (posttest). Because the tDCS intervention consisted of 20 treatment sessions while WBRT consisted of 10 fractions, post-intervention assessments for both groups were performed after completion of the final tDCS session. During the first 10 treatment days, each tDCS session was administered immediately before the corresponding daily WBRT fraction. This study was designed as an exploratory randomized controlled trial. No formal a priori sample size calculation was performed because no previous randomized clinical studies evaluating tDCS during WBRT were available to provide a reliable estimate of the expected effect size. Participants Twenty men and 20 women with a mean age of 57.6 years (standard deviation: 7.8 years; range: 42–81 years) were included in this study. Eligible participants were adults with a confirmed diagnosis of brain metastases who were referred for WBRT at Shahid Madani and Valiasr Hospitals in Tabriz, Iran. All participants were right-handed, able to understand and sign informed consent, and free of concurrent neurological or psychiatric disorders that could confound cognitive outcomes. In this study, all patients receiving WBRT had brain metastases and were considered unsuitable for stereotactic radiosurgery (SRS) following multidisciplinary evaluation by the treating radiation oncologist. Factors contributing to this decision included extensive intracranial disease burden (e.g., more than 15 brain metastases), large cumulative tumor volume, lesions adjacent to critical radiosensitive structures (e.g., the optic nerves, chiasma, or brainstem), or large metastatic lesions (>3–4 cm) for which SRS was considered technically unsuitable. Patients with primary brain tumors were excluded because the study was designed to evaluate cognitive outcomes in patients with metastatic brain disease. In addition, individuals with previous cranial surgery or prior brain radiotherapy were excluded to minimize potential confounding effects on neurocognitive outcomes. Patients receiving medications known to substantially influence cognitive performance or cortical excitability (e.g., cholinesterase inhibitors, memantine, psychostimulants, or other centrally acting cognitive-enhancing drugs) were excluded. Corticosteroid therapy was maintained at a stable dose throughout the study, whenever clinically feasible, to minimize potential confounding effects on cognitive outcomes. Table 1 summarizes the baseline demographic, clinical, and radiotherapy characteristics of participants in the active tDCS and sham groups. Table 1Baseline demographic, clinical, and radiotherapy characteristics of participants in the active tDCS and sham groups.VariableActive tDCS (n = 20)Sham (n = 20)Age (years), mean ± SD57.2 ± 7.458.0 ± 8.2Sex (Male/Female)10 / 1010 / 10Primary diagnosisLung (9)Breast (6)Colorectal (3)Other (2)Lung (8)Breast (8)Colon/rectum (3)Other (1)Total WBRT dose (Gy)30 Gy30 GyNumber of fractions1010Treatment duration (weeks)22Previous non-cranial surgery15 (75%)14 (70%)Previous chemotherapy12 (60%)11 (55%)Previous immunotherapy3 (15%)2 (10%)Corticosteroid use20 (100%)20 (100%)Number of brain metastases, median (range)18 (13–27)17 (15–28)Largest lesion diameter (cm), mean ± SD2.4 ± 1.12.5 ± 1.0Total intracranial tumor volume (cm³), mean ± SD24.8 ± 9.625.5 ± 10.2Total intracranial tumor volume, median (range), cm³22.5 (12–48)23.1 (11–52)Karnofsky Performance Status (KPS), mean ± SD74.2 ± 9.175.6 ± 8.3KPS category, n (%)< 70: 5 (25%)70–80: 9 (45%)≥ 80: 6 (30%)< 70: 4 (20%)70–80: 10 (50%)80 < : 6 (30%) Baseline demographic, clinical, and radiotherapy characteristics of participants in the active tDCS and sham groups. Radiotherapy and tDCS protocols WBRT was delivered using a TomoTherapy system (Accuray Inc., USA) with a 6-MV non-flat photon beam using a helical delivery technique. Patients were immobilized with a thermoplastic head mask, and treatment-planning CT images with a slice thickness of 3 mm were acquired before treatment planning. Treatment parameters, including the total dose and fractionation schedule, are presented in Table 1. All patients received a total dose of 30 Gy to the whole brain (including the cerebral hemispheres, cerebellum, and brainstem) in 10 fractions, as recommended by the NCCN (National Comprehensive Cancer Network) guidelines (Nabors et al., 2023). The tDCS stimulation was delivered using a Segal Stem device (Farmed Tajhiz Co., Iran) at a constant current of 2 mA for 20 min over 20 consecutive weekdays through saline-soaked sponge electrodes (5 × 5 cm; electrode area = 25 cm², current density = 0.08 mA/cm²). Electrode placement followed the international 10–20 EEG system. Each treatment session consisted of two consecutive 10-minute stimulation blocks. During the first block, the anode was positioned over the left dorsolateral prefrontal cortex (F3) and the cathode over the right supraorbital region (Fp2). During the second block, the anode was repositioned over the right dorsolateral prefrontal cortex (F4) and the cathode over the left supraorbital region (Fp1). For each 10-minute montage, the current was ramped up over 30 s, maintained at 2 mA for 9 min, and ramped down during the final 30 s. Sham stimulation used the identical electrode montages and ramp-up/ramp-down procedures but delivered no current during the intervening stimulation period to preserve participant blinding. Sequential bilateral stimulation was selected to modulate cortical excitability in both dorsolateral prefrontal cortices because bilateral prefrontal networks contribute to attention, executive function, and working memory (Keshvari et al., 2013). Each tDCS session was administered immediately before the corresponding daily WBRT fraction during the first 10 treatment days. Participants received a total of 20 weekday tDCS sessions. The remaining 10 sessions were delivered on weekdays after completion of WBRT. All stimulation sessions were conducted by a trained technician in a quiet, controlled environment. Participants were monitored throughout each stimulation session for commonly reported tDCS-related adverse events, including itching, tingling, skin erythema, burning sensation, headache, dizziness, fatigue, nausea, mood changes, and sleep disturbances. Mild transient itching, tingling, skin erythema beneath the electrodes, and occasional headache were the most frequently reported adverse events. No participant discontinued treatment because of adverse events, and no seizures or other serious adverse events occurred during the study. Adverse events were recorded by structured oral questioning after each session; however, no standardized adverse-event questionnaire was used. Assessment tools The continuous performance test (CPT) and a computerized 2-back task (as reported by NCALab) were administered to assess sustained attention, response inhibition, working memory accuracy, and reaction time. In the CPT, participants responded to target letters under two conditions: the X-task and the AX-task. In the X-task condition, participants were instructed to press the response key whenever the letter “X” appeared, regardless of the preceding stimulus. This version primarily assessed vigilance and sustained attention. In the AX-task condition, participants were required to respond to the letter “X” only when it was immediately preceded by the letter “A”. This condition introduced a working memory load by requiring participants to maintain the preceding stimulus in memory and inhibit responses when “X” followed a non-“A” letter. For both conditions, correct detections (hits), omission errors (failures to respond to targets), and commission errors (false alarms to non-targets) were recorded, providing indices of attentional control and inhibitory processing. The primary outcome measures included the number of correct responses (hits), omission errors, and commission errors. Although reaction time was recorded during testing, only accuracy-based measures were included in the present analyses because they represented the predefined cognitive outcomes of interest. In the computerized 2-back task, participants viewed a continuous stream of single-digit numbers (0–9) and were instructed to mentally update and maintain the two most recent stimuli. On each trial, they were required to calculate the sum of the current digit and the one presented two trials earlier, indicating the correct result via keypress. This paradigm imposed high demands on working memory updating, monitoring, and manipulation. Performance on the N-back task was quantified using the number of correct responses, omission errors, and commission errors generated by the computerized testing software. These values represent the task performance scores provided by the software rather than percentage accuracy. All participants completed identical computerized versions of the cognitive tasks at baseline and post-intervention under standardized testing conditions. Resting-state QEEG was recorded using the Segal Stem EEG system (Farmed Tajhiz Company, Tehran, Iran) configured according to the international 10–20 electrode placement system. Electrode impedance was maintained below 5 kΩ, and signals were sampled at 1000 Hz. Each participant underwent 3 min of eyes-open and 5 min of eyes-closed recording. Artifact removal (ocular and muscle) and spectral analysis were performed using NeuroGuide software (v2.8.1, Applied Neuroscience, Inc., St. Petersburg, USA). Regional analyses were performed using predefined frontal (Fp1, Fp2, F3, F4, F7, F8) and temporal (T3, T4, T5, T6, F7, F8) electrode regions. Raw absolute power (μV²) in the theta (4–8 Hz) and beta (13–30 Hz) frequency bands was calculated using Fast Fourier Transform (FFT). Beta activity was analyzed as a single frequency band (13–30 Hz) and was not subdivided into low- and high-beta components. The average raw absolute power across the predefined frontal and temporal regions was entered into the statistical analyses. Normative Z-score maps generated by NeuroGuide were used only for visual illustration and were not included in the statistical analyses. Because statistical analyses were performed on predefined regional averages rather than on individual electrode-wise values, correction for multiple comparisons across electrodes was not required. Western blot analysis Western blot analysis was performed to assess the expression levels of DYRK1A (a member of the dual-specificity tyrosine phosphorylation-regulated kinase), MMP2 (matrix metallopeptidase 2), MMP9 (matrix metallopeptidase 9), VEGF (vascular endothelial growth factor), and β-actin (as the reference for normalization of other protein expression levels) in patients’ blood samples. All participants (n = 40; 20 active tDCS and 20 sham) contributed peripheral blood samples, and all samples were included in the biomarker analyses. Peripheral blood mononuclear cells (PBMCs) were isolated from 10 mL EDTA-treated whole blood by density-gradient centrifugation using Ficoll-Paque (GE Healthcare) according to the manufacturer’s instructions. PBMC pellets were washed with PBS and lysed in RIPA buffer containing protease inhibitors (Complete Mini, Roche). Lysates were centrifuged at 12,000 × g for 10 min at 4 °C, and the supernatants were stored at −80 °C. Protein concentrations were determined using the bicinchoninic acid (BCA) assay (Thermo Fisher). Subsequently, 40 μg of protein per sample was mixed with loading buffer, denatured at 95 °C for 5 min, and separated by 10% SDS-PAGE at 120 V. Proteins were transferred by wet electroblotting using transfer buffer (25 mM Tris, 192 mM glycine, and 20% methanol) at 120 V for 90 min. Membranes were stained with Ponceau S to verify protein transfer and then blocked with 5% non-fat dry milk in TBST for 1 h at room temperature before overnight incubation with primary antibodies at 4 °C (Table 2). After incubation with the primary antibodies, the membranes were washed three times with TBST (20 mM Tris-HCl, 150 mM NaCl, 0.1% Tween−20; 10 min per wash) and then incubated for 1 h at room temperature with the appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies. HRP-conjugated goat anti-rabbit IgG (Cell Signaling Technology, Cat. No. 7074, 1:5000) was used for rabbit primary antibodies (DYRK1A, MMP2, MMP9, and VEGF), whereas HRP-conjugated horse anti-mouse IgG (Cell Signaling Technology, Cat. No. 7076, 1:5000) was used for β-actin. Following secondary antibody incubation, the membranes were washed three additional times with TBST before chemiluminescent detection. Table 2 provides the manufacturer, catalog number, host species, and working dilution for each primary antibody used in this study. Table 2The manufacturer, catalog number, host species, and working dilution for each primary antibody used in this study.TargetSupplierCatalog No.HostDilutionDYRK1AAbcamAb154790Rabbit1:1000MMP2Cell Signaling Technology87809Rabbit1:1000MMP9Abcamab76003Rabbit1:1000VEGFAbcamab46154Rabbit1:1000β-actinSanta Cruz Biotechnologysc−47778Mouse1:1000 The manufacturer, catalog number, host species, and working dilution for each primary antibody used in this study. Protein bands were detected using an enhanced chemiluminescence (ECL) substrate (Cat. No. KGP1121; Nanjing KeyGen Biotech Co., Ltd.) and exposed to X-ray film. Films were scanned at high resolution prior to densitometric analysis. Samples from both study groups were processed under identical experimental conditions, and densitometric analyses were performed using identical image-processing parameters. Band intensities were quantified using ImageJ software (NIH, USA) and normalized to β-actin as an internal control. All samples from each participant were analyzed in duplicate, and densitometric measurements were averaged before statistical analysis. Statistical analysis All statistical analyses were performed using SPSS software (version 26; IBM Corp., Armonk, NY, USA). Descriptive statistics are presented as mean ± standard deviation (SD) for all assessed parameters. Statistical significance was defined as p < 0.05. All statistical tests were two-sided. Normality was assessed using the Shapiro–Wilk test, and homogeneity of variance was evaluated using Levene's test. Baseline demographic and clinical characteristics were compared between groups. For continuous variables, independent-samples t-tests were used, whereas categorical variables were compared using the χ² test. Baseline comparisons showed no significant differences between the active tDCS and sham groups in demographic or clinical characteristics. Separate multivariate analyses of covariance (MANCOVAs) were performed for the cognitive outcome measures. One MANCOVA included the Continuous Performance Test (CPT) correct responses and omission errors as dependent variables, while a second MANCOVA included the N-back correct responses, omission errors, and commission errors. Significant multivariate effects were followed by univariate analyses of covariance (ANCOVAs) for each individual outcome. Absolute beta- and theta-band power were analyzed separately using univariate ANCOVA models. For each analysis, the corresponding baseline value was entered as a covariate to control for pre-intervention performance. Age, sex, tumor diagnosis, corticosteroid use, and radiotherapy parameters were not included as additional covariates because of the limited sample size and exploratory design of the study. No missing data occurred, and all randomized participants completed the study protocol; therefore, all available data were included in the analyses. Because all randomized participants completed the study and no outcome data were missing, the intention-to-treat and per-protocol populations were identical. A post hoc power analysis was conducted using G*Power version 3.1.9.7 to estimate the achieved statistical power of the univariate ANCOVA analyses. The analysis was based on a fixed-effects ANCOVA model with two groups, one baseline covariate, a total sample size of 40, and a two-tailed significance level of α = 0.05. Effect sizes were converted from the observed partial eta-squared values using the formula f=ηp2/(1−ηp2). The observed effect sizes ranged from partial η² = 0.18 for theta-band absolute power to partial η² = 0.4 for N-back correct responses. The estimated achieved statistical power ranged from approximately 0.79 for the smallest observed effect to greater than 0.99 for the largest effects. Because these estimates were based on the observed data, they should be interpreted as descriptive rather than as justification for the original sample size. Results Descriptive findings Representative normative QEEG Z-score maps from one participant in each study group are provided in Supplementary Figure S1 for illustrative purposes only. These maps were not included in the statistical analyses. Group-level conclusions are based on quantitative analyses of raw absolute beta- and theta-band power obtained from predefined electrode regions across all participants. Cognitive and QEEG findings Fig. 1 summarizes the pre- and post-intervention mean (± SD) values for N-back accuracy, CPT performance, and absolute β- and θ-band power of the intervention (tDCS, n = 20) and sham control (n = 20) groups. At baseline, no significant differences were observed between the sham and active tDCS groups for any assessed parameter (p > 0.05). At the post-intervention assessment, the active tDCS group showed clear improvements in cognitive performance compared with the sham group. N-back correct responses increased from 2.96 ± 1.53–4.75 ± 0.55 (about 60.47% improvement). N-back omission errors decreased from 6 ± 0.91–5.15 ± 0.48, and commission errors decreased from 1.05 ± 1.09–0.1 ± 0.30. CPT correct responses increased from 13.20 ± 2.52–18.80 ± 1.39 (about 42.42% improvement). Regarding QEEG measures, absolute beta power decreased from 17.41 ± 4.30–11.46 ± 3.15 μV² (about 51.19% reduction), and theta power decreased from 33.41 ± 5.20–18.15 ± 4.50 μV² (about 84.07% reduction) following active stimulation. Mean ± SD values for cognitive and QEEG outcome measures in the active tDCS and sham groups at pretest (A and B) and posttest (C and D). Panel A shows the cognitive outcomes at pretest (N-back correct responses, omission errors, and commission errors; CPT correct responses and omission errors). Panel B shows pretest beta- and theta-band absolute power (μV²). Panel C shows the corresponding cognitive outcomes at posttest, and Panel D shows posttest beta- and theta-band absolute power (μV²). Table 3, Table 4 summarize the results of the multivariate and univariate ANCOVA analyses and the pairwise comparisons of adjusted means. Table 3Summary of multivariate and univariate ANCOVA results, showing F values, degrees of freedom (df), P-values, partial eta-squared (η²; effect size), and observed statistical power for the effects of tDCS compared with the sham group in patients receiving whole-brain radiotherapy.VariableTestStatisticFdfp-valuePartial η²Observed PowerCPT (Correct responses, Omission errors)MANCOVAWilks' λ = 0.5117.482,36< 0.0010.490.97CPT Correct ResponsesANCOVA—21.641,37< 0.0010.370.99CPT Omission ErrorsANCOVA—13.921,370.0010.270.95N-back (Correct responses, Omission errors, Commission errors)MANCOVAWilks' λ = 0.4315.423,35< 0.0010.570.98N-back Correct ResponsesANCOVA—24.811,37< 0.0010.400.99N-back Omission ErrorsANCOVA—12.081,370.0010.250.93N-back Commission ErrorsANCOVA—9.461,370.0040.200.85Beta Absolute PowerANCOVA—18.371,37< 0.0010.330.98Theta Absolute PowerANCOVA—8.121,370.0070.180.79 Summary of multivariate and univariate ANCOVA results, showing F values, degrees of freedom (df), P-values, partial eta-squared (η²; effect size), and observed statistical power for the effects of tDCS compared with the sham group in patients receiving whole-brain radiotherapy. Table 4Pairwise comparisons of adjusted means between the active tDCS and sham groups for neurocognitive and EEG outcome measures in patients undergoing whole-brain radiotherapy.VariableActive tDCS GroupSham GroupMean Difference (Δ)Std. Errorp-valueN-back Correct Responses5.522.173.350.36< 0.001N-back Omission Errors4.876.28–1.410.45< 0.001N-back Commission Errors–0.091.24–1.330.21< 0.001CPT Correct Responses18.7813.215.570.45< 0.001CPT Omission Errors1.216.78–5.570.45< 0.001Beta Absolute Power (μV²)9.5018.26–8.751.86< 0.001Theta Absolute Power (μV²)14.2022.29–8.092.470.002 Pairwise comparisons of adjusted means between the active tDCS and sham groups for neurocognitive and EEG outcome measures in patients undergoing whole-brain radiotherapy. Baseline demographic and clinical characteristics did not differ significantly between the active tDCS and sham groups (all p > 0.05). Multivariate analysis demonstrated a significant overall effect of treatment on CPT performance (Wilks' λ = 0.51, F(2,36) = 17.48, p < 0.001, partial η² = 0.49) and N-back performance (Wilks' λ = 0.43, F(3,35) = 15.42, p < 0.001, partial η² = 0.57). Follow-up ANCOVA analyses showed that participants receiving active tDCS achieved significantly higher correct response scores and significantly fewer omission and commission errors than those receiving sham stimulation. Significant treatment effects were observed for CPT correct responses (F(1,37) = 21.64, p < 0.001), CPT omission errors (F(1,37) = 13.92, p = 0.001), N-back correct responses (F(1,37) = 24.81, p < 0.001), N-back omission errors (F(1,37) = 12.08, p = 0.001), and N-back commission errors (F(1,37) = 9.46, p = 0.004). ANCOVA also demonstrated significantly lower beta-band absolute power (F(1,37) = 18.37, p < 0.001) and theta-band absolute power (F(1,37) = 8.12, p = 0.007) in the active tDCS group compared with the sham group after adjustment for baseline values. Pairwise comparisons of adjusted means showed that participants receiving active tDCS exhibited significantly better performance than those receiving sham stimulation on both cognitive tasks. Adjusted mean correct-response scores were significantly higher in the active tDCS group, whereas omission and commission error scores were significantly lower (all p < 0.001). Similarly, adjusted beta- and theta-band absolute power values were significantly lower in the active tDCS group than in the sham group (both p < 0.01). Biomarker (western blot) analysis results Representative band intensity profiles and mean ± standard deviation (error bars) from the Western blot analysis of DYRK1A, MMP2, VEGF, MMP9, and β-actin before and after the intervention are shown in Fig. 2, Fig. 3. Representative immunoblots are presented in Fig. 2, whereas quantitative analyses were performed using normalized band intensities obtained from all participants. The corresponding uncropped Western blot images are provided in Supplementary Figure S2. Representative Western blot images of DYRK1A, MMP2, VEGF, MMP9, and β-actin in the active tDCS group, obtained before treatment and after completion of the intervention period. Molecular weight markers are shown (kDa) in parenthesis below each protein label. The corresponding uncropped Western blot images are provided in Supplementary Fig. S2. Mean ± SD of normalized protein expression levels for DYRK1A (a), MMP2 (b), VEGF (c), and MMP9 (d) in peripheral blood samples collected before and after the intervention in the active tDCS and sham groups. Protein expression was normalized to β-actin. * indicates a statistically significant difference between the active tDCS and sham groups at post-intervention (p < 0.05), whereas # indicates a statistically significant difference from the corresponding pre-treatment value within the same group (p < 0.05). Molecular weight markers (kDa) and lane labels are shown for the representative immunoblots. Exact p-values are reported in the results. As shown in Fig. 3, the expression levels of DYRK1A, MMP2, VEGF, and MMP9 increased following WBRT in the sham group. In contrast, participants receiving active tDCS showed significantly lower post-intervention expression levels of these proteins than the sham group, together with significant reductions compared with their own pre-treatment values. These findings indicate that active tDCS was associated with reduced peripheral expression of DYRK1A, MMP2, VEGF, and MMP9 following WBRT. Detailed exploratory correlation analyses are provided in Supplementary Table S1. Safety outcomes No serious adverse events related to active tDCS or sham stimulation were observed throughout the study period. The intervention was well tolerated by all participants, and no participant discontinued treatment because of adverse effects. Mild transient adverse events were reported in both groups. In the active tDCS group, five participants reported headache, seven developed mild erythema beneath the stimulation electrodes, seven experienced transient itching, and six reported transient tingling. In the sham group, six participants reported headache, two experienced transient itching, and three reported transient tingling. All adverse events resolved spontaneously without medical intervention, and no seizures or other clinically significant complications attributable to tDCS were observed during or immediately after the stimulation sessions. Discussion This study evaluated the effects of tDCS and QEEG markers (beta- and theta-band absolute power) on working memory, sustained attention, blood levels of DYRK1A, MMP2, VEGF, and MMP9, and EEG markers in brain tumor patients undergoing WBRT. Compared with sham stimulation, active tDCS significantly improved cognitive test scores, reduced peripheral expression of DYRK1A, MMP2, VEGF, and MMP9, and reduced beta- and theta-band absolute power on QEEG. These findings suggest that tDCS may provide cognitive and neurophysiological benefits during WBRT. However, the underlying biological mechanisms remain to be established. A previous study by Roy et al (Roy et al., 2025). reported that tDCS over the ventromedial prefrontal cortex (vmPFC) mitigated stress-induced impairments in working memory, reduced physiological stress responses, increased correct responses, and decreased omission errors in the CPT. Consistent with these findings, we observed that tDCS significantly improved working memory and sustained attention. In patients with brain tumors, who often experience attention deficits due to neuroinflammation or direct brain injury, tDCS has been shown to enhance performance, possibly by modulating frontoparietal networks involved in attention control (van Grinsven et al., 2021, Zhang et al., 2025, Stagg and Nitsche, 2011). Improvements in working memory, as measured by the N-back task, may reflect improved function of prefrontal and frontoparietal networks (Pupíková et al., 2021, Jones et al., 2020). Previous studies have reported that WBRT can impair neuronal structure and reduce synchronization across distributed networks (Zhang et al., 2018, Mitchell et al., 2020 Feb 1, Warrington et al., 2013). These findings are consistent with the possibility that tDCS influences functional connectivity within networks supporting working memory, although functional connectivity was not directly measured in the present study (Zaehle et al., 2011, Claaß et al., 2024). Our results regarding the reduction in absolute beta-band power following active tDCS contrast with several prior studies reporting increases in beta-band power during heightened cognitive engagement and attentional control (Wen et al., 2024, Palacios-García et al., 2021). While beta synchronization has been associated with top-down modulation and sustained alertness, recent evidence suggests that excessive beta activity in prefrontal regions may result from compensatory hyperexcitability or inefficient cortical processing (Bressler and Richter, 2015). The observed reduction in beta-band power may reflect more efficient cortical oscillatory processing following tDCS. However, because cortical excitability and network efficiency were not directly measured, this interpretation should be considered hypothetical (Yang et al., 2015). The significant reduction in resting-state theta-band absolute power observed in the active tDCS group may indicate partial normalization of abnormal slow wave activity associated with brain metastases and WBRT. Increased resting-state theta activity has been reported in several neurological disorders and has been associated with diffuse cortical dysfunction, cognitive fatigue, and impaired cognitive performance (Dörre, 2023). The observed reduction in theta power following tDCS may therefore reflect changes in cortical oscillatory activity accompanying the improvements in working memory and sustained attention observed in the present study. Although previous investigations have demonstrated that hippocampal–prefrontal theta synchronization plays an important role in working memory encoding and maintenance (Soltani Zangbar et al., 2020 May 1), the present study evaluated resting-state EEG rather than task-related oscillatory activity. Therefore, these findings should not be interpreted as evidence of altered hippocampal–prefrontal theta synchrony during working memory. These findings are consistent with tDCS-induced modulation of frontoparietal theta–gamma coupling reported in working memory training studies (Jones et al., 2020, Ikeda et al., 2019). Our results also align with Nissim et al (Nissim et al., 2023)., who demonstrated that transcranial alternating current stimulation (tACS) can enhance working memory performance in healthy individuals. Although the present study employed tDCS rather than tACS, the similarity in modulating theta-band power may account for the observed cognitive improvement. Furthermore, several reviews and experimental studies have shown that tDCS enhances cortical excitability and can improve cognitive performance in neurological populations (Zaehle et al., 2011, Medeiros et al., 2012). Exploratory correlation analyses demonstrated modest associations between reductions in beta- and theta-band power and improvements in cognitive performance (Supplementary Table S1). However, because these analyses were exploratory and the study was not powered for mediation analyses, no causal relationships can be inferred. Therefore, it cannot be concluded that the electrophysiological changes directly mediated the observed cognitive improvements. Future studies incorporating correlation or mediation analyses are needed to clarify these relationships. DYRK1A is a kinase involved in multiple cellular processes, including cell cycle regulation, apoptosis, and neurogenesis. Brain radiotherapy can induce DNA damage, cell death, and inflammatory responses, and DYRK1A expression has been found to increase in response to cellular stress and DNA damage, likely as a compensatory mechanism (Rammohan et al., 2022). Evidence from CRISPR and mechanistic studies indicates that DYRK1A influences DNA-repair pathways and apoptosis, and its modulation can alter radiosensitivity in cancer cell models (Lan et al., 2022). tDCS could decrease DYRK1A levels in peripheral blood, via ablation of compensatory processes following cellular damage. However, because DYRK1A was measured in peripheral blood mononuclear cells, the present study cannot determine whether similar changes occurred within the central nervous system. DYRK1A's role as a transcriptional regulator is more recently being recognized. Furthermore, DYRK1A appears to be overexpressed following radiotherapy in primary brain tumors like glioblastoma (Kim et al., 2021, Liu et al., 2021 Sep). Plasma DYRK1A levels and their temporal characterization post-irradiation, has not been well characterized. Radiosensitivity of related DYRK family members (e.g., DYRK3) has also been reported to be upregulated in glioblastoma models, upon irradiation (Kim et al., 2021). This supports a concept that stress-responsive DYRK family kinases can be radiation inducible. MMP2 and MMP9 are enzymes involved in extracellular matrix (ECM) degradation and tissue remodeling. They are often upregulated in response to cellular injury, inflammation, and tumor invasion (Wolosowicz et al., 2024), and play a key role in the reparative process by facilitating the migration of cells to sites requiring repair. In gliomas and other brain tumors, elevated MMP2 and MMP9 levels are associated with increased invasiveness and poorer prognosis, with experimental studies linking them to enhanced tumor cell migration and angiogenesis (Pullen et al., 2018 Sep). Preclinical studies indicate that cranial irradiation can disrupt the MMP2/TIMP2 axis, increase MMP expression, and contribute to blood–brain barrier (BBB) disruption and post-radiation tissue remodeling (Lee et al., 2012). Multiple animal and early human studies have demonstrated anti-inflammatory effects following repeated tDCS—including reductions in pro-inflammatory cytokines and microglial activation—supporting an immunomodulatory role of tDCS that could plausibly reduce MMP induction (Ethridge et al., 2022). Furthermore, tDCS may promote neurogenesis and tissue repair in a more regulated manner, potentially contributing to lower peripheral MMP2 and MMP9 expression. Nevertheless, these findings should be interpreted cautiously because the present study did not directly assess neuroinflammation, blood–brain barrier integrity, or extracellular matrix remodeling. We also assessed VEGF concentrations in peripheral blood of sham and active tDCS groups’ participants before and after radiotherapy. VEGF is the major hypoxia-inducible pro-angiogenic factor that is responsible for endothelial cell proliferation and neovessel formation (Green et al., 2007). Radiation injury results in microvascular damage and tissue hypoxia, and the concomitant activation of hypoxia-inducible factor 1 (HIF−1) – and, in many models, a direct effect of radiation – results in up-regulated VEGF expression and angiogenesis, contributing to the post-irradiation vasculature response, which may be conducive to tumor regrowth (Moeller et al., 2004, Kim et al., 2004, Telarovic et al., 2021 Dec). Hypoxia-induced VEGF expression is an adaptive process which leads to formation of new blood vessels and restoration of oxygen and nutrients to damaged tissue, but may also be conducive to tumor growth, providing the tissue with blood supply for further growth and spread. tDCS has been shown to modulate neurovascular regulation, which could lead to improved tissue oxygenation and hypoxia reduction, and therefore suppression of the pro-angiogenic stimulus (Cancel et al., 2022, Ryu et al., 2025, Ren et al., 2024). By modulating neurovascular regulation, tDCS may obviate the need for angiogenesis, therefore resulting in a lower VEGF expression. Previous experimental studies have suggested that tDCS may influence neurovascular and inflammatory processes. The reduced peripheral VEGF expression observed in the present study may therefore be consistent with altered systemic biological responses following tDCS. However, tissue hypoxia, angiogenesis, and cerebral blood flow were not directly measured, and these potential mechanisms remain speculative. The observed improvements in working memory and sustained attention may have clinical relevance, as cognitive impairment is among the most disabling consequences of WBRT. Nevertheless, the present study was not designed to determine the minimum clinically important difference, and future studies should evaluate whether these improvements translate into meaningful benefits in functional independence and quality of life. In addition, because patients undergoing WBRT differ from other neurological populations with respect to tumor-related pathology, and concurrent medical treatments, the mechanisms and magnitude of response to tDCS may not be directly comparable to those reported in healthy individuals or other clinical populations. This study has several limitations. The sample size was relatively modest, all patient data were collected from a single center, and no a priori sample size calculation was performed. Consequently, the study may have been underpowered to detect smaller treatment effects or to perform meaningful subgroup analyses according to tumor characteristics or treatment variables. Although baseline demographic, clinical, and treatment characteristics were summarized in Table 1, some clinically relevant variables, including baseline neurological deficits, baseline cognitive status, and concomitant medications other than corticosteroids, were not systematically collected. Therefore, residual clinical heterogeneity cannot be excluded and should be considered when interpreting the findings. The study did not include a healthy control group. Although normative Z-score maps were used for visual reference, future studies including healthy controls could help distinguish radiotherapy-specific effects from general cognitive changes associated with tDCS. In addition, repeated administration of cognitive tests may have introduced practice effects. Although both groups completed the same testing protocol and the sham-controlled design partially accounts for improvements related to repeated testing, practice effects cannot be completely excluded. Future studies should consider using alternate task versions and longer follow-up intervals to minimize this potential source of bias. Furthermore, the biomarker analyses were exploratory and based on peripheral blood measurements; therefore, these findings should be interpreted cautiously until replicated in larger studies with more comprehensive molecular characterization. Because outcomes were assessed immediately after completion of the intervention, the durability of the observed cognitive and electrophysiological improvements remains unknown. Finally, blinding success was not formally evaluated, and future studies should also incorporate validated safety assessment instruments to provide a more comprehensive evaluation of treatment tolerability. Larger adequately powered multicenter studies with longer follow-up are needed to confirm the efficacy, durability, safety, and underlying mechanisms of tDCS in patients undergoing WBRT. Conclusion Active tDCS administered immediately before each daily WBRT session and continued for 10 additional sessions after completion of WBRT was associated with short-term improvements in working memory and sustained attention, accompanied by reductions in beta- and theta-band absolute power on QEEG and lower peripheral expression of DYRK1A, MMP2, VEGF, and MMP9. These findings suggest that tDCS may have beneficial short-term cognitive and neurophysiological effects throughout the treatment period; however, the underlying biological mechanisms remain to be established. Given the exploratory nature of the biomarker analyses, the modest sample size, and the absence of long-term follow-up, larger adequately powered multicenter studies are required to confirm the efficacy, durability, safety, and underlying mechanisms of tDCS in patients undergoing WBRT. CRediT authorship contribution statement Davood Khezerloo: Writing – original draft, Visualization, Validation, Conceptualization. Hassan Sabouri Moghaddam: Writing – review & editing, Visualization, Validation, Supervision, Project administration, Methodology, Formal analysis, Data curation. Touraj Hashemi: Writing – review & editing, Formal analysis, Data curation. Leila Halavi Anjar: Writing – original draft, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Amir Ghasemi Jangjoo: Validation, Supervision, Project administration, Methodology, Investigation. Consent for publication Not applicable. Ethics approval and consent to participate This study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board (IRB) of Tabriz University (approval code: IRTABRIZU.TEC.1403.107). Written informed consent was obtained from all participants prior to enrollment, and all data were anonymized to preserve confidentiality. No serious adverse events related to tDCS were reported; mild side effects (e.g., itching, tingling, or erythema) were recorded at each session. Declaration of Generative AI and AI-assisted technologies in the writing process The authors declare that the artificial intelligence-based tools were applied to only do language editing and grammar and clarity improvements. The authors take full responsibility to the integrity and honesty of the presented material. Funding None. Clinical Trial Registration Not applicable. Declaration of Competing Interest The authors declare that there is no conflict of interest regarding the publication of this article.