Onkologikoa

Polyphenols represent the largest and most diverse class of dietary antioxidants. Epidemiological evidence linking specific (poly)phenol classes, such as flavonoids and lignans, to breast cancer (BC) risk remains limited and largely inconclusive in prospective studies. The aim of this study is to examine the association between the intake of total (poly)phenols-and its classes and subclasses-and BC risk-overall and by subtypes (estrogen, progesterone, and human epidermal growth factor receptor 2 (HER2))-in the European Prospective Investigation into Cancer and Nutrition (EPIC) cohort. The EPIC cohort includes 257,960 adult women from seven European countries. During a mean follow-up of 14 years, there were 10,722 incident overall BC cases. Associations were computed using Cox regression models adjusted for potential confounders. No significant associations were found between total (poly)phenol intake and overall BC risk (HRQ5 vs. Q1 = 1.02; 95% CI: 0.95-1.11). In addition, null associations were mostly found between classes and subclasses of (poly)phenols and BC subtypes. After stratifying by menopausal status, no significant associations were observed. In conclusion, this study found no evidence of associations between the intake of any class or subclass of (poly)phenols and BC risk in the European population.

Abstract Brain tumours invade neural tissue, disrupting the functional organisation of neural networks. This disruption can trigger compensatory neuroplastic mechanisms that help preserve cognitive function despite pathological burden. Resting-state functional connectivity (rs-FC) provides a valuable approach for examining these alterations, yet the scope and trajectory of network-level changes remain unclear. In this study, we used rs-FC to characterise compensatory responses induced by left-hemisphere brain tumours affecting the integrity of the language network. The comparison of connectivity patterns at diagnosis and after surgery allowed us to track the temporal progression of these patterns. Relative to healthy participants, patients showed significant deviations in pre-surgical rs-FC, with widespread changes involving the dorsal attention, sensorimotor, frontoparietal, default mode, and cerebellar networks. These differences persisted post-surgically, with no significant longitudinal modifications. Network-level alterations also influenced topological properties: patients showed increased segregation while preserving global integration relative to controls. These properties changed significantly from pre- to postoperative period, with a postoperative increase in segregation and integration towards a pattern more closely resembling that of healthy controls. Furthermore, preoperative network segregation was identified as a predictor of postoperative cognitive recovery. These findings suggest that brain tumours induce network remapping, affecting functional connectivity far beyond the regions strictly related to the affected region. The observed changes in network topological organisation likely reflect the adaptive role of domain-general networks in facilitating homeostatic processes. Surgical strategies aimed at preserving these changes seem to promote sustained cognitive function, regardless of malignancy grade or lesion extent. Key points Presurgical patients show widespread inter- and intrahemispheric rs-FC alterations, reflecting tumour-driven network remapping. Network integration and segregation show longitudinal recovery, approaching neurotypical topology four months post-surgery. Preoperative network segregation predicts postoperative cognitive recovery. Importance of the Study Brain tumours disrupt large-scale brain networks, yet the extent, temporal dynamics, and cognitive relevance of these alterations remain poorly understood. Using resting-state functional connectivity, this study provides a longitudinal, network-level characterisation of tumour-induced neuroplasticity in patients with left-hemisphere lesions affecting language-related regions. Compared with prior work focused on local effects or single networks, our results demonstrate widespread inter- and intrahemispheric connectivity changes involving multiple domain-general resting-state networks, highlighting extensive network remapping beyond lesion boundaries. Importantly, we show that postoperative recovery is accompanied by a rebalancing of network integration and segregation toward a neurotypical configuration. Critically, preoperative network segregation emerges as a predictor of postoperative cognitive recovery, identifying a potential biomarker of resilience. These findings have direct translational relevance, suggesting that preserving and supporting large-scale network organisation during surgical planning may promote cognitive outcomes independently of tumour grade or lesion extent.

Over the past few years, several fluid biomarker candidates have been proposed for frontotemporal dementia (FTD). We have previously identified CSF proteins that could separate individuals with genetic FTD from controls. However, it is unknown whether alterations in these CSF protein levels are associated with neurodegenerative processes. The aim of this study was to explore how these CSF biomarker candidates correlate with symptom severity as well as cortical and subcortical atrophy. The levels of fourteen proteins were measured in CSF from 202 individuals, 131 mutation carriers with mutations in C9orf72, GRN, or MAPT, and 71 controls, in a cross-sectional subset from the GENFI cohort. The association between the levels of these proteins and CDR plus NACC FTLD-NM sum-of-boxes, cortical thickness, and subcortical volumes were estimated in the mutation carriers. Elevated CSF levels of five out of fourteen proteins were associated with an increased CDR score in the mutation carriers. Additionally, elevated levels of three of these proteins, NEFM, PTPRN2 and SERPINA3, were associated with reduced cortical thickness and/or subcortical volume among all mutation carriers. Some mutation-specific associations were also observed, with SPP1 and CTSS being associated with CDR and atrophy only in MAPT mutation carriers, while NPTX2 was specific for GRN mutation carriers. As indicated by the association to brain atrophy, the proposed fluid biomarker candidates continue to show promise and additional studies will further elucidate their relationship to cortical atrophy in genetic FTD, and their potential as biomarkers for diagnosis, prognosis, and disease staging.

Abstract Inflammation is a natural immune response that can protect against cancer in early stages but, when becomes chronic, creates an immunosuppressive tumour microenvironment (TME) that promotes tumour progression. Recognized as a hallmark of cancer, chronic inflammation offers potential for new therapeutic strategies, particularly immunotherapy. However, some solid tumours, such as breast cancer, show partial resistance to these treatments. Cytokines are key regulators of inflammation, influencing cell communication and immune responses within the TME (1,2). Importantly, current clinical classification of BC often neglects the immune landscape, limiting the ability to predict responses to immunotherapy. Our hypothesis is that the cytokine profile of metastatic breast cancer possesses prognostic and predictive value and that cytokines can be potential biomarkers in breast cancer. Understanding how cytokine loops are coordinated within the breast TME could also lead to novel cytokine-based therapeutic strategies. Cytokine levels in serum and fresh tissue samples from metastatic BC patients (n=150) and in serum samples from metastatic cancer patients from other solid tumour types (n=186) were measured by Olink cytokine panel, which quantitatively assess 40 cytokines. Cytokine levels were correlated with clinical data, including overall and relapse free survival. The statistical analysis was done using different mathematical modelling methods. Our unpublished results show that the serum and tumour cytokine profiles do not correlate significantly, supporting the idea that some cytokines act locally in the TME. In addition, comparison with other metastatic tumour samples, led to the identification of some cytokines that are differentially expressed in BC. Integration of these data will shed light on the characterization of cytokine interactions in metastatic BC, helping to define its immune landscape. 1. Soler MF, Abaurrea A, Azcoaga P, Araujo AM, Caffarel MM. New perspectives in cancer immunotherapy: targeting IL-6 cytokine family. J Immunother Cancer. 2023;11(11):e007530; 2. Manzano S, Caffarel MM. Cytokine-centered strategies to boost cancer immunotherapy. Mol Oncol. 2025;19(3):579-583 Citation Format: M. M. CaffarelU. AlvarezP. AzcoagaA. RomoS. ManzanoP. PetitI. GarmendiaM. GarzónZ. TelletxeaM. OtañoA. UrruticoecheaI. Alvarez-Lopez. Characterization of the cytokine profile of metastatic Breast Cancer, from mediators of chronic inflammation to potential biomarkers [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS4-03-16.

N6-methyladenosine (m6A) RNA modifications and reovirus infections have recently been implicated in the development of celiac disease. While viral infections are known to alter the RNA modification machinery of the host, their interplay in autoimmune tissue damage remains unexplored. In this study we employed an in vitro model combining a viral mimic with gliadin peptides to simulate celiac disease conditions, alongside the analysis of serum and intestinal biopsies from controls and patients to investigate the link between m6A methylation and viral infections in inducing autoimmune inflammation. We found elevated anti-reovirus reactivity in patients, increased antiviral gene expression and enhanced m6A levels. Notably, the expression of IRF7 was synergistically induced by the combined exposure to the viral mimic and gliadin peptides, and this is mediated through m6A methylation in its coding region and interaction with the reader protein YTHDC2. Furthermore, the reduction of m6A through METTL3 silencing or simvastatin treatment reduced IRF7 mRNA methylation and downstream proinflammatory gene expression both in vitro and ex vivo. These findings highlight m6A methylation as a modulator of antiviral responses and a potential therapeutic target in autoimmune disorders.

BackgroundThe left atrium (LA) and left atrial appendage (LAA) are common sites of thrombus formation leading to cardioembolic stroke. We aimed to include LA/LAA in acute-phase stroke CT angiography (CTA) to detect potential embolic sources early.MethodsThe CTA scan range was extended to include the LA/LAA. We retrospectively reviewed CTAs from patients treated at Donostia University Hospital between December 2020 and November 2021. LA/LAA images were classified as normal filling, non-well-defined filling defect, and thrombus-suggestive filling defect. Clinical, demographic, and stroke severity data were compared across groups.ResultsAmong 299 patients included, 6 (2%) had thrombus-suggestive filling defects, and 32 (11%) had non-well-defined defects. Both groups had more severe strokes (median NIHSS 16 vs. 13, p < 0.01), higher mortality (29% vs. 6%, p < 0.01), and more pre-existing heart disease (74% vs. 27%, p < 0.01) and known atrial fibrillation (AF) (63% vs. 8%, p < 0.01). AF was more frequently diagnosed during follow-up in the non-well-defined filling defect group than in the normal group (50% vs. 19%, p = 0.02). Prior AF was the only independent predictor of thrombus-suggestive filling defect, which was observed despite high rates of anticoagulation (50%).ConclusionsIncluding LA/LAA in acute CTA allows early identification of thrombus-suggestive and non-well-defined filling defects, both linked to worse outcomes and potential therapeutic consequences.Supplementary InformationThe online version contains supplementary material available at 10.1007/s00415-026-13692-6.

Atogepant is approved for migraine prevention and has shown strong efficacy in clinical trials. However, its effectiveness following failure of anti-CGRP monoclonal antibodies (MAbs) has not been evaluated in large real-world populations. This multicenter observational study conducted across Spanish headache units included adults with migraine who initiated atogepant after failure of ≥ 1 anti-CGRP MAb and had ≥ 3 months of follow-up. Baseline demographic and clinical variables were collected prospectively, with follow-up assessments at months 3 and 6. The primary outcome was the proportion of patients achieving a ≥ 50% reduction in monthly migraine days (MMD) at three months. Secondary outcomes included ≥ 30%, ≥ 75%, and 100% response rates; changes in headache days, pain intensity, acute medication use, and patient-reported outcomes; adverse events; treatment persistence; and factors associated with response. A total of 252 patients were included (mean age 48.9 ± 12 years; 83.3% female; 80.6% with chronic migraine; 45.6% with continuous daily headache). Prior to atogepant, 39.7% had failed one anti-CGRP MAb, 27.0% two, 20.2% three, and 13.1% four. Median baseline MMD was 16, monthly headache days 27, and acute medication days 20. At 3 months, 44.4% achieved a ≥ 30% reduction in MMD, 29.7% ≥50%, and 11.7% ≥75%. Adverse events were reported in 52.5% of patients, most commonly constipation (30%) and nausea (25%). At three months, 26.2% had discontinued treatment (65.1% due to inefficacy, 28.8% due to intolerance). Treatment persistence at 180 days was 61% (95% CI 54 to 69%). A higher number of previously failed MAbs was independently associated with reduced odds of ≥ 50% response (RR 0.79, 95% CI 0.64 to 0.97). Moreover, a higher number of previously failed MAbs was associated with diminished improvements across multiple clinical endpoints, including headache frequency, intensity, acute medication use, and disability measures. Atogepant may represent a viable treatment option for patients with migraine who have failed anti-CGRP MAbs. In this large real-world cohort, approximately one-third of patients achieved a ≥ 50% response, despite a treatment-refractory profile. However, the likelihood of response decreases with a higher number of previously failed MAbs, and mild adverse events are frequent. This is the largest real-world study to date evaluating atogepant after failure of anti-CGRP monoclonal antibodies in migraine, including a highly refractory multicenter cohort. Atogepant after anti-CGRP monoclonal antibodies failure achieved a ≥ 50% response in 29.7% of patients and a ≥ 30% response in 44.4% at 3 months, with mild adverse effects reported in over half of cases. A higher number of previously failed MAbs, trigeminoautonomic symptoms, and depression were independently associated with lower likelihood of response.