GGD Amsterdam

Background & Aims Up to 15% of ulcerative colitis (UC) patients with an ileo‐anal pouch will develop chronic or antibiotic refractory pouchitis. We aimed to evaluate the efficacy and safety of tofacitinib in these refractory patients. Methods In this prospective single center study, adult UC patients with chronic or antibiotic refractory pouchitis (pouchitis disease activity index [PDAI] ≥ 7) were treated with tofacitinib 10 mg twice daily for 8 weeks. Clinical, biochemical, endoscopic and histologic disease activity was assessed at baseline and at week 8. The primary endpoint was the proportion of patients achieving clinical remission (PDAI < 7 and a reduction of ≥3 points from baseline) or clinical response (decrease ≥2 points from baseline). Results Thirteen patients were included and treated with tofacitinib. After 8 weeks, 31% achieved PDAI‐defined remission (4/13) and 54% achieved response (7/13). Both the total PDAI score (11 [interquartile range 9–12.75] vs. 8 [4.5–9.75], p = 0.033) and the clinical PDAI subscore (4 [3–4] vs. 2 [0.25–3.75], p = 0.014) decreased significantly from baseline compared to week 8 or early withdrawal, respectively. We did not observe a change in endoscopic or histological PDAI subscores. Conclusion In this pilot study, clinical remission in patients with chronic pouchitis was achieved in 31% of patients after 8 weeks of treatment with tofacitinib. Total and clinical PDAI dropped significantly compared with baseline, but we did not observe a significant change in endoscopic or histologic disease activity. Presumably, a treatment duration of 8 weeks is insufficient to induce mucosal healing in these refractory patients.

448 Background: T-DXd is a HER2-directed antibody-drug conjugate; T-DXd 6.4 mg/kg monotherapy is approved for patients with metastatic HER2+ GC/GEJA who have received a prior trastuzumab-based regimen. Preliminary results from DG-03 Part 2 showed feasibility and promising antitumor activity with 1L T-DXd 6.4 mg/kg (arm D) or 5.4 mg/kg (arm F) with FP and pembrolizumab in patients with esophageal adenocarcinoma/GC/GEJA. T-DXd 6.4 mg/kg and FP with pembrolizumab was associated with higher than expected toxicity; alternatively, early safety data from T-DXd 5.4 mg/kg triplet combination showed a manageable safety profile. We report updated results for arm F from DG-03 Part 2 with an approximate time-matched analysis with arm D. Methods: DG-03 (NCT04379596) is a Phase 1b/2 multicenter, open-label, dose-escalation (Part 1) and -expansion (Parts 2, 3, and 4) study. In Part 2, patients with HER2+ (immunohistochemistry [IHC] 3+ or IHC 2+/in situ hybridization–positive by local testing) esophageal adenocarcinoma/GC/GEJA, irrespective of programmed cell death ligand 1 status, and no prior treatment for metastatic disease were enrolled. In arm D, patients received T-DXd 6.4 mg/kg intravenous (IV) infusion every 3 weeks (Q3W) and FP (5-fluorouracil [5-FU] 600 mg/m ² continuous IV or capecitabine [cape] 1000 mg/m ² twice daily [BID]) with pembrolizumab 200 mg IV Q3W; in arm F, patients received T-DXd 5.4 mg/kg IV Q3W and FP (5-FU 600 mg/m ² continuous IV or cape 750 mg/m ² BID) with pembrolizumab 200 mg IV Q3W.Primary endpoint was confirmed objective response rate by investigator assessment (INV) per RECIST 1.1. Secondary endpoints included duration of response by INV and progression-free survival by INV. Safety and tolerability were also assessed. Results: A time matched analysis was performed to compare the efficacy of arm D (n=43), data cut-off (DCO) October 27, 2022, and arm F (n=32), DCO May 6, 2024, given the limited follow up of the 5.4mg/kg. The median duration of follow up for the two cohorts was 4.1 months and 4.6 months for arm D and arm F, respectively. Objective response rate for arm D was 41.9% and 59.4% for arm F. Median PFS was 6.4 months (95% CI: 5.0, NC) for arm D and 5.8 months (95% CI: 5.6, NE) for arm F. At DCO, the median overall survival in both arms was not reached. Updated time matched safety and efficacy data will be provided at the time of presentation. Conclusion: The time matched analysis shows that lowering the dose of T-DXd to 5.4 mg/kg from 6.4 mg/kg and lowering the starting dose of capecitabine, did not result in decrease in efficacy of the combination of T-DXd + FP + pembrolizumab. Funding: This study is sponsored by AstraZeneca in collaboration with Daiichi Sankyo. In March 2019, AstraZeneca entered into a global development and commercialization collaboration agreement with Daiichi Sankyo for trastuzumab deruxtecan (T-DXd; DS-8201). Editorial acknowledgment: Under guidance of the authors and in accordance with Good Publication Practice, medical writing and editorial support was provided by Carmen Grimaldos, PhD, of Helios Medical Communications, part of Helios Global Group, and was funded by AstraZeneca. Clinical trial information: NCT04379596 .

Introduction: Early identification of the etiology of ischemic stroke is crucial for secondary prevention. Recent publications on dedicated ECG-gated cardiac CT showcase the potential to detect sources of cardiac embolism. Photon-counting CT (PCCT) uses semiconductors to convert X-ray photons into electrical signals directly. This allows for improved image quality and lower radiation dose compared to conventional CT. Exploiting PCCT technology with a high-pitch dual-source image acquisition may render an exceptionally high temporal resolution. We hypothesized that non-ECG-gated high-pitch dual-source PCCT angiography from the diaphragm to the brain might provide added clinical value in detecting cardiac stroke sources during initial stroke imaging while maintaining optimal brain and neck image quality. Methods: Consecutive patients with a clinical suspicion of acute stroke imaged with a PCCT system between October 4 th , 2023 and April 13 th , 2024 at a Swedish comprehensive stroke center were included. Diaphragm to brain coverage was obtained in an acquisition time of 1.3 seconds; on average, yielding a dose-length product of 360 mGy*cm for the study participants (2.34 mSv). Image quality was graded using a 4-point Likert scale. Images were assessed for cardiac stroke sources. Where available, reference standard echocardiography results were collected. Results: Of 249 consecutive stroke investigations, 193 included cardiothoracic imaging, which constituted the study population. 126 scans (65.3%) were cases with imaging confirmed ischemic stroke. The median age was 74 (IQR, 61-81) years, 99 were male (51.3%). In total, 39.4% underwent follow-up echocardiography. Image quality of the heart was excellent in 19 scans (9.8%) good in 97 (50.3%), moderate in 72 (37.3%), and poor in 5 scans (2.6%). Fourteen (7.3%) certain or probable interatrial septal defects were found in the study population. In the imaging-confirmed ischemic stroke subgroup, six certain or probable cardiac thrombi were found (4.8%). Additionally, a possible cardiac thrombus was found in 31 instances (24.6%). Three aortic valve vegetations were found (2.4%), all confirmed by echocardiography. Conclusions: Non-ECG-gated high-pitch dual-source PCCT angiography typically provides images of high quality, at virtually no time expense while maintaining a reasonably low radiation exposure. Non-gated PCCT angiography is a promising technique to be used as a primary screening method for cardioembolic stroke.

We propose a method for extracting the axial length of the human eye from high-resolution spectral domain optical coherence tomography (SD-OCT) retinal scans. The method evaluates the chromatic dispersion introduced by the anterior segment and the vitreous of the eye. By analyzing sub-spectral scans, we quantify the axial shift caused by dispersion and relate it to the thicknesses of the media passed by the OCT beam. The method depends on accurate k-linearization and pixel-to-wavenumber calibration. First, we demonstrate the feasibility of our approach using a model eye with adjustable water chamber length. Subsequently, the method is explored for in vivo retinal OCT scans. Challenges are inter-subject variability and limited availability of exact chromatic dispersion data for ocular tissues in the relevant spectral range for OCT imaging. By interpolating the refractive indices of an established eye model from visible wavelengths to the infrared wavelengths of the OCT system using the dispersion of water and estimating refined dispersion properties of the lens, we improve the model’s agreement with in vivo measurements.

INTRODUCTION Blood‐based glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL), and phosphorylated tau (pTau) have shown promising prognostic potential in Alzheimer's disease (AD), but their applicability in clinical settings where comorbidities are prevalent remains uncertain. METHODS Simoa assays quantified GFAP, NfL, and pTau181 in retrospectively retrieved prediagnostic serum samples from 102 AD patients and 21 non‐AD controls. RESULTS Higher serum GFAP levels predicted earlier clinical presentation and faster subsequent Mini‐Mental State Examination decline in AD patients. Serum NfL levels were increased in patients with arterial hypertension (AHT), kidney dysfunction, and a history of stroke and only demonstrated predictive value for time to clinical AD presentation after adjustment for these comorbidities. Serum pTau181 instability during long‐term storage at −20°C prevented its prognostic evaluation in retrospectively retrieved serum samples. DISCUSSION Serum GFAP is a robust prognostic marker for AD progression, whereas NfL is impacted by various comorbidities, which complicates the interpretation of its prognostic value. Highlights Serum GFAP levels predict time to clinical AD presentation. Serum NfL levels are increased by hypertension, kidney disease, and stroke history. Prognostic value of serum NfL in AD is only evident after comorbidity correction. Serum levels of GFAP, but not NfL, increase over time within prediagnostic AD stages.

Background Inadequate sleep among Dutch adolescents is a complex public health issue with detrimental effects on physical and mental well-being. Previous interventions have shown limited or no lasting effects. While systems approaches help to understand and address such challenges, effective prevention efforts require evidence-based intervention design using behavior change techniques. This study outlines the systematic development of a ‘whole systems action program’ , named Charge Your Brainzzz (CYB), to promote healthy sleep in Dutch adolescents aged 12-15, by integrating theory- and evidence-based behavior change methods using the Intervention Mapping Protocol into a systems science approach, combined with stakeholder engagement Methods The CYB program was developed based on previously identified key system dynamics influencing adolescent sleep and a detailed action plan targeting these dynamics. The development was guided by five steps inspired by the Intervention Mapping protocol: defining the program goal and prioritizing system dynamics outcomes across various interconnected subsystems (step 1), specifying performance objectives (step 2), constructing matrices of change objectives (step 3), developing program components while selecting theoretical methods and practical applications (step 4), and program production and formative testing (step 5). Adolescents, parents, teachers, school boards, school care coordinators, youth healthcare professionals, and Healthy School advisors were actively involved. Results The whole systems action program CYB consists of eight program components and includes: 1) an educational component, 2) a step-by-step guide for implementing school sleep health policies, 3) a parent information evening, 4) an online magazine, 5) Teen Sleep Check, 6) Sleep Guide, 7) Tool for monitoring and early detection for sleep (behavior) problems, and 8) implementation materials including a website. Conclusions The CYB program is the first whole systems action program designed to promote adolescent sleep health. Using a Systems Science approach, the Intervention Mapping (IM) protocol, and stakeholder engagement, the program addresses the complexity of the health issue, the lived experiences of the target groups, and a solid theoretical foundation for behavior change. Beyond presenting the program’s components, this study offers a replicable roadmap for addressing complex public health challenges, paving the way for innovative, system-oriented solutions in health promotion.

Objective To assess the pathological mechanisms contributing to white matter (WM) lesion expansion or contraction and remyelination in multiple sclerosis (MS). Methods We assessed 1,613 lesions in 49 people with relapsing–remitting MS in the CCMR‐One bexarotene trial (EudraCT 2014‐003145‐99). We measured lesion orientation relative to WM tracts, surface‐in gradients and veins. Jacobian deformation was used to assess lesion expansion over 6 months, while magnetization transfer ratio (MTR) imaging was used to assess remyelination. Results At baseline, 33% of lesions were aligned with veins, 2% along WM tracts, 0% with surface‐in gradients, and 4% orthogonal to veins. No significant differences were observed in lesion shape, while lesions aligned with surface‐in gradients and with veins had lower volume compared to all remaining orientations. At follow‐up, 13% of lesions expanded and 7% contracted. The directions for both expansion and contraction were 18% and 8%, respectively, along WM tracts, 20% and 15% parallel to veins, 22% and 23% orthogonal to veins and 0% and 1% along surface‐in gradients. Bexarotene had no effect on lesion expansion or contraction, but MTR significantly increased in lesions aligned with surface‐in gradients and veins. Interpretation Lesion expansion and shrinkage are affected by venous and WM tract factors, but these do not influence bexarotene's capacity to promote remyelination. This, instead, appears to be affected by surface‐in factors. To limit lesion expansion and maximize tissue repair, multiple processes may need to be targeted.

Background. A clinical prediction model (IMPROVE) for ipsilateral ischemic stroke risk in symptomatic patients with carotid artery disease was recently developed and internally validated with good performance. In this decision-analytic study, we evaluated the model-based cost-effectiveness of IMPROVE-based triage versus triage in care-as-usual (CAU) for optimal medical treatment (OMT) alone or carotid revascularization plus OMT. Methods. A dataset of 678 patients with a recent ipsilateral ischemic stroke, TIA or amaurosis fugax pooled from 4 cohort studies, informed a decision-analytic model. Stratification of patients for carotid revascularization was either based on ≥50% carotid stenosis (CAU arm), or based on a range of 3-year ipsilateral ischemic stroke risk thresholds (IMPROVE arm). The threshold resulting in the lowest number of strokes was selected as the optimal threshold. Patients with <50% stenosis (CAU arm) or an IMPROVE risk score below the threshold were modeled to receive OMT only. Parameter uncertainty was incorporated in probabilistic analyses using 10,000 Monte Carlo simulations for a 3-year and lifetime horizon. Subgroup analyses for mild (<50%), moderate (50-69%), and severe (70-99%) carotid stenosis were performed. Results. IMPROVE-based triage, as compared to triage in CAU, reduced ipsilateral ischemic strokes and perioperative stroke/death by 34.5% (CAU: 4.3% vs. IMPROVE: 2.8%) over 3 years. Perioperative stroke occurred in 1.8% of the patients in CAU versus 1.4% for IMPROVE. Revascularizations decreased by 20% with IMPROVE, while QALYs slightly increased. Societal costs decreased on average by €1,441/patient for IMPROVE versus CAU for a 3-years time horizon (lifetime cost reduction: €6,101/patient). Subgroup analyses identified IMPROVE as the superior strategy for moderate/ severe stenosis (3-year and lifetime horizon) and mild stenosis (lifetime horizon). Conclusions. Triage of symptomatic patients with carotid artery disease with the IMPROVE model versus CAU can lead to the prevention of one-third of ipsilateral ischemic strokes, while also reducing societal costs.

Introduction The ethical obligation to reduce the environmental impact of healthcare systems prompts an exploration of if and when environmental concerns should be integrated into clinical decision‐making. In this study, we aimed to elucidate patients' attitudes regarding the provision of environmental information in healthcare decision‐making and to identify preferred approaches for integrating these considerations into patient–provider consultations. Methods This interview study served as an in‐depth follow‐up of a survey study on gynaecological patients' perspectives on environmental sustainability within healthcare settings. We conducted semistructured interviews with 14 patients from two Dutch outpatient clinics between February and May 2024. We employed reflexive thematic analysis to analyse the data. Results Five main themes were developed from the data: (1) Patients are an integral part of the transition to sustainable healthcare, (2) Patients are open to information on environmental impact of healthcare, (3) Information on environmental impact should be tailored to the individual patient and context, (4) Patients vary in preferences for involvement in decision‐making related to environmental sustainability and (5) Patients prioritize individual health over environmental concerns in healthcare decision‐making. Conclusion The findings of our study underscore the importance of integrating sustainability into clinical decision‐making, aligning with bioethical principles and the expectations and goals of patients. By ensuring that environmental considerations are introduced in a personalized and context‐appropriate manner within patient–provider interactions, healthcare can foster greater support for sustainable practices. Patient or Public Contribution Patients were involved in developing the pre‐interview questionnaire. The preliminary results of the study were presented to healthcare professionals from various backgrounds during a meeting of the Dutch Green Care Alliance, after which the input was incorporated into the interpretation of the study results. Finally, the complete manuscript was presented to representatives from the Dutch Patient Federation to obtain their input on the interpretation and implications of our research. The interpretation of our results aligned with the findings from their recent national inquiries, which included data from over 9300 patients within their network. Consequently, no changes were made to the manuscript's discussion.

Shallow genome-wide cell-free DNA (cfDNA) sequencing holds great promise for non-invasive cancer monitoring by providing reliable copy number alteration (CNA) and fragmentomic profiles. Single nucleotide variations (SNVs) are, however, much harder to identify with low sequencing depth due to sequencing errors. Here we present Nanopore Rolling Circle Amplification (RCA)-enhanced Consensus Sequencing (NanoRCS), which leverages RCA and consensus calling based on genome-wide long-read nanopore sequencing to enable simultaneous multimodal tumor fraction estimation through SNVs, CNAs, and fragmentomics. Efficacy of NanoRCS is tested on 18 cancer patient samples and seven healthy controls, demonstrating its ability to reliably detect tumor fractions as low as 0.24%. In vitro experiments confirm that SNV measurements are essential for detecting tumor fractions below 3%. NanoRCS provides the opportunity for cost-effective and rapid processing, which aligns well with clinical needs, particularly in settings where quick and accurate cancer monitoring is essential for personalized treatment strategies.

Differentiation of antigen-activated B cells into pro-proliferative germinal center (GC) B cells depends on the activity of the transcription factors MYC and BCL6, and the epigenetic writers DOT1L and EZH2. GCB-like Diffuse Large B Cell Lymphomas (GCB-DLBCLs) arise from GCB cells and closely resemble their cell of origin. Given the dependency of GCB cells on DOT1L and EZH2, we investigated the role of these epigenetic regulators in GCB-DLBCLs and observed that GCB-DLBCLs synergistically depend on the combined activity of DOT1L and EZH2. Mechanistically, inhibiting both enzymes led to enhanced derepression of PRC2 target genes compared to EZH2 single treatment, along with the upregulation of BCL6 target genes and suppression of MYC target genes. The sum of all these alterations results in a 'cell identity crisis', wherein GCB-DLBCLs lose their pro-proliferative GC identity and partially undergo PC differentiation, a state associated with poor survival. In support of this model, combined epi-drugging of DOT1L and EZH2 prohibited the outgrowth of human GCB-DLBCL xenografts in vivo. We conclude that the malignant behavior of GCB-DLBCLs strongly depends on DOT1L and EZH2 and that combined targeting of both epigenetic writers may provide an alternative differentiation-based treatment modality for GCB-DLBCL.

Background In a 16‐week, 91‐patient placebo‐controlled clinical study in DLB (“AscenD‐LB”;NCT04001517), neflamapimod improved outcomes on the CDR Sum‐of‐Boxes (p = 0.023 vs. placebo) and Timed Up and Go test (p = 0.044 vs. placebo). Further, at the higher of two dose levels in the study neflamapimod also improved (p = 0.049 vs. placebo) outcomes in a cognitive test battery, particularly with respect to attention (p = 0.023 vs. placebo). The results were most prominent in patients with pure DLB (i.e., without Alzheimer’s disease‐related co‐pathology, assessed by plasma ptau181). While no validated biomarkers for treatment effects in DLB were available when the study was conducted, with a recent report that plasma GFAP and neurofilament‐light‐chain (NfL) discriminated MCI‐DLB from healthy controls, with GFAP being more discriminant, we evaluated the effects of neflamapimod in AscenD‐LB on plasma GFAP and NfL, both overall and in pure DLB patients. Method GFAP and NfL levels (pg/mL) in stored plasma samples from AscenD‐LB were determined using Simoa® platform. Design and overall results (Jiang et.al. Nature Communications, 2022), and analysis after stratification for baseline plasma ptau181 (Alam et.al. Neurology, 2023) are published. Result At baseline, GFAP was higher (p = 0.02) in DLB with AD co‐pathology [282(SD = 120, n = 29] versus in pure DLB [215(SD = 91), n = 28]; NfL was not. In the overall population, a reduction in GFAP levels were seen in neflamapimod‐treated participants (n = 30, ‐12.3±6.8, baseline‐to‐week16) and not in placebo‐recipients (n = 27, +3.7±7.8; p = 0.13); p = 0.13 for the difference. In pure DLB participants, the change from baseline to week 16 in GFAP was significantly different (p = 0.04) in neflamapimod‐treated participants (n = 15, ‐10.6±6.4) compared to placebo‐recipients (n = 13, +14.1±10.2). For NfL, no neflamapimod‐placebo differences were evident in either population. In the pure DLB population, the change in GFAP levels from baseline was significantly correlated to change in CDR‐SB (r = 0.54, p = 0.036; increased GFAP associated with worsening CDR‐SB, reduction in GFAP associated with improvement on CDR‐SB) among neflamapimod recipients, but not in placebo‐recipients. Conclusion The beneficial effects on GFAP further support that neflamapimod is clinically efficacious in patients with DLB. The results also demonstrate the potential of GFAP as biomarker to assess treatment effects in DLB, though treatment effects may be masked in patients with AD co‐pathology.

Background Irsenontrine (e2027) is a potent and selective PDE9 inhibitor that increases cellular cGMP which is important for glutamatergic synaptic function. Irsenontrine was investigated to improve cognition in Lewy Body Dementia (LBD; DLB and PDD), and recent phase 2 study data suggests that irsenontrine could be more effective in DLB patients without amyloid copathology. Here, we evaluated differential change from baseline levels in proteins associated with cGMP pathway in DLB participants without amyloid co‐pathology (DLB A‐) compared to DLB participants with amyloid co‐pathology (DLB A+). We next evaluated the affected biological pathways. Method We employed proximity extension analysis (PEA) proteomics in all available paired pre‐ and post‐treatment CSF samples of a translational medicine trial of E2027 (Figure 1A). This set included six DLB A‐ patients (CSF Lumipulse Ab42/40 cut‐off 0.057) and nine DLB CSF A+ patients with a mean±sd age of 76±6 years and 67% male. We assessed the treatment effect per DLB group by paired Wilcoxon tests of pre‐ vs post‐treatment (1) protein levels (p<0.05) and (2) individual pathway enrichment z‐scores based on 189 KEGG pathways (p<0.05). Result Following E2027 treatment, CSF protein changes were apparent in both groups (DLB A‐ 10 proteins increased, and 32 decreased; DLB A+ 16 proteins increased, and 27 decreased; Figure 1B/C). We found that more pathways were affected in DLB A‐ (5 pathways), compared to DLB A+ (3 pathways), and several of these were related to the mechanism of action, including an increase in Pantothenate and CoA biosynthesis z‐scores in both groups, and lower GABAergic synapse in DLB A‐ only (Figure 1D/E). Conclusion This global proteomics analyses in DLB clinical trial samples presents that, despite the small number of participants treated with E2027, there were changes in proteins related to the mechanism of action, as indicated by Pantothenate and CoA biosynthesis increase, which amongst others factors related to glutamate metabolism. In addition, multiple proteins and pathways were differentially affected between the DLB A‐ and A+ group, which could lead to improved understanding of differential clinical treatment response in DLB patients.

Background The Amyloid Imaging to Prevent Alzheimer’s Disease (AMYPAD) Prognostic & Natural History Study (PNHS) is a prospective longitudinal PET cohort of over 1,500 non‐demented individuals from 10 parent cohorts across Europe. We provide an overview of ongoing efforts to curate and integrate magnetic resonance imaging (MRI) multimodal images across sites and to extract biologically meaningful information (i.e., image‐derived phenotypes; IDPs) in this early AD population. Data will be made available on the ADDI platform (fair.addi.ad‐datainitiative.org). Method Imaging protocols included core (T1w, T2w, T2*, FLAIR) and advanced (rs‐fMRI, SWI, DWI, and ASL) MRI sequences. Figure 1 provides an overview of the different acquisitions, their corresponding pipelines, and the obtained IDPs. For T1w images, we computed regional volumes and thickness values using FreeSurfer v7.1.1 (surfer.nmr.mgh.harvard.edu/). Morphometric similarity networks were computed using MIND (github.com/isebenius/MIND). White matter hyperintensity volumes were obtained from FLAIR images using Bayesian Model Selection (BaMoS). With rs‐fMRI, mean functional connectivity of canonical networks were extracted using FSL MELODIC and dual‐regression analyses after preprocessing with fMRIPrep v23.0.1 (fmriprep.org). Diffusion MRI scans were processed with QSIprep v0.16 (qsiprep.readthedocs.io) to compute tract‐based spatial statistics and tractograms to build structural connectivity matrices. Finally, cerebral blood flow and the spatial coefficient of variation were computed from ASL images using ExploreASL (github.com/ExploreASL). Statistical harmonization was performed using RELIEF (github.com/junjypark/RELIEF). Result Baseline and follow‐up characteristics of the 1537 subjects with available T1w are described in Table 1. Raw and statistically harmonized variants of 380 core and 298 advanced IDPs were calculated. Figure 2 shows the distribution of hippocampal volumes per site before and after statistical harmonization, and their relation with participants' age (R2raw=0.179, R2harmonized=0.422). Considering recent developments in the field, the AMYPAD PNHS aims to include additional IDPs associated with vascular health, such as perivascular spaces and other glymphatic system‐related markers. Conclusion We present the pipeline built for MRI harmonization and feature extraction of the AMYPAD PNHS dataset. The extracted IDPs can help identify novel imaging outcomes, support the development of disease progression models for the preclinical stages of AD, and provide a reference point for future studies to promote replicability and robustness of findings.

Background Different patterns of atrophy exist in the dementia stage of AD. However, little is known about the heterogeneity of atrophy patterns and the mechanisms that drive subsequent propagation of the disease in the preclinical stages. Method From the AMYPAD‐PNHS cohort, we included a total of 1323 non‐demented individuals, including 1094 amyloid‐negative, and 229 amyloid‐positive participants (Table 1). Gray matter thickness in 100 regions of interest was measured using FreeSurfer. Global cortical amyloid burden was derived from amyloid PET scans in Centiloids (CL). Individuals were assigned to atrophy subtypes based on cortical thickness measures using Non‐Negative Matrix Factorization (Figure 1). Multinomial regression models were used to study differences between subtypes in age, sex, global CL and APOE E4 carriership. Linear regression models were used to investigate the effect of time on atrophy within each subtype. Coordinated deformation models (Figure 1, Shafiei et al. 2023) were used to estimate network‐based thickness changes in each region, by multiplying the effect of time in connected regions by the strengths of their connections. Connectivity strength was based on templates of functional, structural and morphometric similarity networks, and allowed us to evaluate differential mechanisms. Significance was tested using a permutation approach. Results Three significant subtypes of regional GM atrophy were identified:‘limbic predominant’, ‘typical AD’, and ‘diffuse cortical’ (Figure 2). Individuals assigned to the limbic predominant subtype were significantly older, while the typical AD subtype had more APOE E4 carriers compared to the others (all p<0.001). Both limbic predominant and typical AD subtypes had higher CL compared to the diffuse cortical atrophy subtype (all p<0.001). Significant longitudinal thickness changes per subtype are shown in Figure 2. In the typical AD subtype, all network architectures constrained longitudinal GM changes (p<0.001), while in the diffuse atrophy subtype, propagation was mostly determined by morphometric similarity (p<0.005). No significant prediction model was found for the limbic predominant subtype. Conclusion We show that atrophy subtypes can already be detected in non‐demented individuals. Progression of atrophy within subtypes is determined by network‐based mechanisms.

Background Sabirnetug (ACU193) is a humanized IgG2 antibody targeting soluble, synaptotoxic amyloid β oligomers (AβOs). AβOs accumulate in Alzheimer’s disease (AD) and induce pre‐ and post‐synaptic changes, resulting in dendritic spine loss, neuronal degeneration, and release of synaptic proteins into the CSF. Recently, we reported that three administrations of sabirnetug in an early AD population (INTERCEPT‐AD Phase 1 study, NCT04931459) significantly lowered CSF levels of the post‐synaptic protein neurogranin as well as pTau181. Here, we present data on additional CSF synaptic biomarkers and AD plasma biomarkers measured in INTERCEPT‐AD. Method INTERCEPT‐AD was a randomized, placebo‐controlled study with two parts: single ascending dose (SAD) randomized in a 6:2 ratio to sabirnetug (2, 10, 25, 60 mg/kg) or placebo and multiple ascending dose (MAD) randomized 8:2 to sabirnetug (three administrations at 10 or 60 mg/kg every 4 weeks [Q4W] or 25 mg/kg Q2W) or placebo. Biomarkers were measured blinded in CSF and EDTA‐plasma, before and after drug administration, by Amsterdam UMC. Result CSF levels of vesicle‐associated membrane protein 2 (VAMP2), a protein involved in pre‐ and post‐synaptic vesicle trafficking, were significantly normalized (lowered) in sabirnetug treated participants relative to placebo in all three MAD cohorts. Levels of the pre‐synaptic protein neuronal pentraxin 2 (NPTX2), which acts on post‐synaptic excitatory synapses, regulates complement activity, and lowers with cognitive decline, trended lower with sabirnetug than placebo. Plasma levels of GFAP, pTau181, and pTau217 trended towards normalization (lowering) of AD‐dependent changes in the 60 mg/kg Q4W cohort. Conclusion In INTERCEPT‐AD, three administrations of sabirnetug lowered CSF levels of both pre‐ and post‐synaptic proteins, consistent with sabirnetug’s proposed mechanism of action to inhibit AβO synaptic binding. VAMP2 appeared most sensitive to sabirnetug in this study, lowering significantly in all three MAD cohorts; other markers previously reported to have a statistically significant response to sabirnetug ‐ neurogranin and pTau181 ‐ did so at the highest dose. No statistically significant changes in plasma biomarkers were observed in this short study. Long‐term changes in biomarker levels and their relationship to clinical efficacy will be evaluated in the 18‐month ALTITUDE‐AD phase 2 study of sabirnetug beginning in the first half of 2024.

Background The Amyloid Imaging to Prevent Alzheimer’s Disease (AMYPAD) Prognostic & Natural History Study (PNHS) is a prospective longitudinal PET cohort of over 1,500 non‐demented individuals from 10 parent cohorts across Europe. We provide an overview of ongoing efforts to curate and integrate magnetic resonance imaging (MRI) multimodal images across sites and to extract biologically meaningful information (i.e., image‐derived phenotypes; IDPs) in this early AD population. Data will be made available on the ADDI platform (fair.addi.ad‐datainitiative.org). Method Imaging protocols included core (T1w, T2w, T2*, FLAIR) and advanced (rs‐fMRI, SWI, DWI, and ASL) MRI sequences. Figure 1 provides an overview of the different acquisitions, their corresponding pipelines, and the obtained IDPs. For T1w images, we computed regional volumes and thickness values using FreeSurfer v7.1.1 (surfer.nmr.mgh.harvard.edu/). Morphometric similarity networks were computed using MIND (github.com/isebenius/MIND). White matter hyperintensity volumes were obtained from FLAIR images using Bayesian Model Selection (BaMoS). With rs‐fMRI, mean functional connectivity of canonical networks were extracted using FSL MELODIC and dual‐regression analyses after preprocessing with fMRIPrep v23.0.1 (fmriprep.org). Diffusion MRI scans were processed with QSIprep v0.16 (qsiprep.readthedocs.io) to compute tract‐based spatial statistics and tractograms to build structural connectivity matrices. Finally, cerebral blood flow and the spatial coefficient of variation were computed from ASL images using ExploreASL (github.com/ExploreASL). Statistical harmonization was performed using RELIEF (github.com/junjypark/RELIEF). Result Baseline and follow‐up characteristics of the 1537 subjects with available T1w are described in Table 1. Raw and statistically harmonized variants of 380 core and 298 advanced IDPs were calculated. Figure 2 shows the distribution of hippocampal volumes per site before and after statistical harmonization, and their relation with participants' age (R² raw=0.179, R² harmonized=0.422). Considering recent developments in the field, the AMYPAD PNHS aims to include additional IDPs associated with vascular health, such as perivascular spaces and other glymphatic system‐related markers. Conclusion We present the pipeline built for MRI harmonization and feature extraction of the AMYPAD PNHS dataset. The extracted IDPs can help identify novel imaging outcomes, support the development of disease progression models for the preclinical stages of AD, and provide a reference point for future studies to promote replicability and robustness of findings.

Background The increasing dementia prevalence and potential introduction of disease‐modifying therapies (DMTs) highlight the need for efficient diagnostic pathways. Clear recommendations to guide the choice of diagnostic tests are lacking and may vary depending on different clinical scenarios. We used a data‐driven approach to identify efficient and effective stepwise diagnostic testing for three clinical scenarios: 1) syndrome diagnosis, 2) etiological diagnosis, 3) potential eligibility for DMT. Method We used data from two memory clinic cohorts (ADC, PredictND), including 504 patients with dementia (302 Alzheimer’s disease, 107 frontotemporal dementia, 35 vascular dementia, 60 dementia with Lewy bodies), 191 patients with mild cognitive impairment, and 188 cognitively healthy controls (CN). Tests included digital cognitive screening (cCOG), neuropsychological and functional assessment (NP), MRI with automated quantification, and CSF biomarkers. Sequential testing followed a predetermined order (Figure 1). Subsequent tests were conducted if the diagnosis remained uncertain. Diagnostic certainty was ascertained through a data‐driven clinical decision support system (CDSS) that generated a disease state index probability score (DSI, 0‐1), indicating the probability of each diagnosis. Diagnosis was confirmed if the DSI exceeded a predefined threshold, set based on sensitivity/specificity cutoffs relevant for each clinical scenario and step. We assessed correct diagnoses and the need for additional testing at each step. Result For syndrome diagnosis, stepwise testing (cCOG, NP, MRI) accurately identified 71% of the patients, with NP needed in 42%, and MRI in 31%. For etiological diagnosis, starting with cognitive testing to rule out dementia resulted in the need for MRI in 84% of cases, including 91% of dementia patients and 25% CN. Subsequent MRI reduced CSF required to 29%, ultimately diagnosing 81% of patients with 71% accuracy. In determining DMT eligibility, stepwise testing (cCOG, NP, MRI) correctly identified 91% of potential eligible patients for confirmatory CSF testing, while only 51% of ineligible patients. Conclusion Depending on the setting, alternative diagnostic pathways are accurate and efficient. As such, a data‐driven tool can assist clinicians in selecting tests of added value across different clinical contexts. This becomes especially important with DMT availability, where the need for more efficient diagnostic pathways is crucial to maintain accessibility and affordability of diagnoses.

Background Plasma levels of amyloid‐β1‐42/1‐40, and glial fibrillary acidic protein (GFAP) have demonstrated predictive potential for amyloid pathology in the early stages of Alzheimer’s disease (AD) development. Utilizing baseline and up to 6‐year follow‐up plasma, positron emission tomography (PET) and cognitive data from cognitively unimpaired individuals, we here aim to test whether early changes in plasma biomarker levels associate with change in amyloid status and cognitive decline. Methods From the EMIF‐AD PreclinAD study we selected individuals with normal cognition and longitudinal plasma, PET and cognitive data available (nbaseline=200, table 1). Amyloid‐status was determined using visual assessment of dynamic [¹⁸F]flutemetamol‐PET scans. Plasma levels of amyloid‐β1‐40, amyloid‐β1‐42, neurofilament light (NfL), and GFAP were measured using the neurology 4‐plex E Simoa assay (Quanterix). Linear mixed models were applied to examine longitudinal changes in plasma biomarker levels, and their association with longitudinal changes in amyloid‐status and cognition. Results At baseline 13.3%, and at 4‐year follow‐up 25.9% of participants were amyloid‐positive. Twenty‐one participants converted from a negative to positive amyloid‐status. At baseline, GFAP levels were higher and amyloid‐β1‐42/1‐40 lower in amyloid‐positive compared to negative individuals (β(SE)=0.43(0.14), β(SE)=‐0.75(0.21), p<0.005, respectively). In the total cohort, plasma levels of GFAP and NfL increased over time, whereas amyloid‐β1‐42/1‐40 remained stable (Figure 1). Although overall stable, plasma amyloid‐β1‐42/1‐40 decreased in participants that converted from a negative to positive amyloid‐status over time (β(SE)=‐0.04(0.02), p=0.02). Alongside, an increase in GFAP levels was observed in participants that converted from a negative to positive amyloid‐status compared to stable negative participants (Figure 1). No group differences were observed for NfL. A longitudinal increase in GFAP levels was associated with memory decline in stable amyloid positive individuals (GFAP*time*amyloid‐group status: β(SE)=‐0.09(0.02), p<0.001, Figure 2). Conclusions Early changes in plasma GFAP and amyloid‐β1‐42/1‐40 are associated with changes in amyloid‐status in cognitively unimpaired individuals. Further, increases in plasma GFAP levels over time are predictive of future memory decline in amyloid‐positive individuals. These results support the potential of plasma GFAP and, to a lesser extent, amyloid‐β1‐42/1‐40 as diagnostic and prognostic markers for AD in early stages of disease development.