Quantitative Measurement of Extracellular Vesicle‐Sized Particles in Blood Plasma by Laser Diffraction Without an Extraction Process: A Proof‐of‐Concept Study in Atrial Fibrillation Authors: Giichi Nitta, Kotone Tatsumi, Nurjahon Ummatalieva, Suzuka Ishikawa, Satomi Hamada, Anna Suzuki, Rina Iwasaki, Kohei Kawajiri, Ryutaro Kobayashi, Junji Yamaguchi, Ryunosuke Ohkawa, Kotaro Yoshioka, Tetsuya Nagata, Takanori Yokota, Kensuke Ihara, Tetsuo Sasano Published: 2026 ABSTRACT ABSTRACTBackgroundExtracellular vesicles (EVs) circulate in blood and may serve as disease biomarkers. However, in atrial fibrillation (AF), reliable circulating biomarkers remain limited. Conventional quantification of EVs requires cumbersome extraction procedures that can reduce quantitative accuracy. We investigated whether laser diffraction (LD) enables accurate and reproducible measurement of EV‐sized particles in blood plasma without an extraction process.MethodsLD was used to measure particle size and concentration of synthetic liposomes and plasma‐derived EV fractions, and the results were compared with nanoparticle tracking analysis (NTA). Particles in plasma were then measured by LD with and without an extraction process, and the impact of dietary conditions was evaluated. Finally, particle concentrations in fasting plasma were compared between 20 controls without AF and 20 patients with AF.ResultsFor both synthetic liposomes and extracted EV fraction, LD showed quantitative accuracy comparable to NTA, with higher reproducibility. LD‐based measurement in plasma without an extraction process produced a size distribution profile similar to that obtained after extraction, whereas the peak particle concentration decreased after extraction, suggesting partial particle loss during processing. Postprandial samples showed contamination by chylomicrons and very‐low‐density lipoproteins, which shifted the size distribution toward larger particles and interfered with plasma measurement. Using fasting plasma, the concentration of EV‐sized particles measured by LD was significantly higher in AF patients than in controls (p = 0.0225).ConclusionsLD enables accurate and highly reproducible quantification of EV‐sized particles in fasting plasma without an extraction process and may support proof‐of‐concept biomarker assessment in AF. Introduction Extracellular vesicles (EVs) are lipid bilayer‐enclosed nanoparticles secreted by many cell types into the extracellular space and play an important role in intracellular communication carrying proteins and nucleic acids [1, 2]. EVs are generally classified into microvesicles, exosomes, and apoptotic bodies based on their biogenesis and size, spanning a broad diameter range from approximately 30 to 1000 nm. Since EVs are released by cardiomyocytes, endothelial cells, and inflammatory cells, growing attention has been directed toward their role in cardiovascular diseases [3, 4]. In particular, circulating EVs have been investigated as potential biomarkers in heart failure, ischemic heart disease, and arrhythmia, such as atrial fibrillation (AF) [5, 6, 7, 8]. AF is the most common sustained cardiac arrhythmia and is associated with increased risks of heart failure, ischemic stroke, and impaired quality of life [9, 10]. Although early detection and accurate diagnosis are essential for initiating appropriate therapies [11, 12]. AF is often asymptomatic or paroxysmal, which complicates timely identification. Consequently, there is growing interest in reliable circulating biomarkers that reflect AF‐related pathophysiology and may support early diagnosis [13, 14]. For circulating EVs to be applied as clinically relevant biomarkers, reliable and reproducible quantification methods are required. Current EVs analyses generally rely on EVs isolation from blood plasma using ultracentrifugation, size‐exclusion chromatography, co‐precipitation, or immunoaffinity‐based techniques [15, 16, 17, 18]. These procedures involve multiple centrifugation and resuspension steps, which may introduce technical variability and limit quantitative accuracy [19, 20]. Importantly, such complex preprocessing hampers the feasibility of EV‐based biomarkers in large‐scale studies and routine clinical settings, including those focused on AF. Nanoparticle Tracking Analysis (NTA) has been the most commonly used method for EVs measurement at the single‐particle level to measure the particle size and concentration of EVs [21, 22]. However, the reproducibility of EVs concentration and size measurements by NTA can be affected by measurement settings, sample heterogeneity, and device‐dependent attributes [23, 24]. Laser Diffraction (LD) is a method to irradiate a laser to a group of particles and detect diffraction and scattering patterns originating from the particles. Recently, the measurement of proteins and silica beads by LD has been reported and the particles with diameters ranging from 200 nm to 1000 nm could be evaluated [25, 26]. However, there have been no reports on the systematic EVs measurement by LD. We hypothesized that LD was applicable for the quantitative measurement of EVs in the blood sample. Thus, we evaluated the accuracy and reproducibility in the measurement of EVs in comparison with NTA as a standard. Furthermore, we evaluated whether this approach could facilitate simple EV quantification in blood samples and explored the potential utility of EVs measured by LD as a biomarker for AF. Methods Synthetic Liposomes Synthetic liposomes (1,2‐dioleoyl‐sn‐glycero‐3‐phosphocholine [DOPC] / Cholesterol [CHOL] (55:45 mol/mol) (FormuMax Scientific Inc.)) of 100 nm in diameter were used to validate the accuracy of the measurements. The same samples of synthetic liposomes were measured by NTA and LD three times. The average particle size of the liposomes was provided by the manufacturer and was 112.1 ± 0.3 nm. Blood Collection and Plasma Separation This study was approved by the institutional review board of Tokyo Medical and Dental University School of Medicine (Approved number: M2018‐140). Informed consent was obtained from each patient and the study protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a prior approval by the institution's human research committee of Tokyo Medical and Dental University School of Medicine. Peripheral blood samples were collected between November 10, 2021 and December 7, 2021 from controls without AF (20 samples) and patients with AF (20 samples) who visited Tokyo Medical and Dental University Hospital. All patients with AF were hospitalized for pulmonary vein isolation (PVI). Blood samples were collected on the day of admission after an overnight fast of at least 8 h, and PVI was performed on the following day. The control group consisted of 14 patients admitted for coronary angiography, 4 patients admitted for electrophysiological study/catheter ablation for paroxysmal supraventricular tachycardia or premature ventricular contractions, and 2 patients admitted for endovascular therapy for arteriosclerosis obliterans. In these control patients, blood samples were collected under conditions similar to those in the AF group, namely on the day of admission after an overnight fast of at least 8 h, and the planned procedure was performed on the following day. 7 mL of blood was collected in a vacuum collection tube containing ethylenediaminetetraacetic acid‐2Na [EDTA‐2Na] from the participants. The collected blood sample was centrifuged at 1300 × g for 10 min at 4°C, and the supernatant was transferred to a clean tube and further centrifuged at 16 000 × g for 10 min at 4°C. The plasma was stored at 4°C and used within 4 h. The plasma was used for LD measurement without dilution. EVs Extraction To extract EVs, Total Exosome Isolation reagent (Thermo Fisher Scientific) was used under the guidance of the manufacturer's protocol. Briefly, 600 μL of plasma was dispensed into a 1.5 mL tube. Then, 300 μL of phosphate buffered saline (PBS) and 180 μL of exosome precipitation reagent were added and incubated at room temperature for 10 min after vortexing. After incubation, the mixture was centrifuged at 10 000 × g for 5 min at room temperature, and the pellet including the EVs was resuspended in 600 μL of PBS. EVs Characterization Resuspended EVs solution was used for LD without dilution and was diluted 100‐fold with PBS for NTA. TEM was performed to confirm the morphology of the isolated EVs using a JEM‐1400Flash transmission electron microscope (JEOL, Tokyo, Japan) operated at an acceleration voltage of 50 kV. For molecular characterization, Western blotting was conducted. The EVs pellets were lysed with RIPA buffer supplemented with protease inhibitors. Equal amounts of protein were separated by SDS‐PAGE and transferred to a polyvinylidene difluoride (PVDF) membrane. After blocking with 5% non‐fat milk in TBS‐T, the membranes were incubated overnight at 4°C with primary antibodies against CD63 (MX‐49.129.5, sc‐5275, Santa Cruz) and Alix (1A12, sc‐53 540, Santa Cruz). Following washing steps, membranes were incubated with HRP‐conjugated secondary antibody (anti‐mouse IgG, #7076, Cell Signaling Technology) for 1 h at room temperature. Protein bands were visualized using Chemi‐Lumi One Super (Nacalai Tesque) for CD63 and Chemi‐Lumi One Ultra (Nacalai Tesque) for Alix, prepared by mixing Solution A and Solution B at a 1:1 ratio. The signals were detected and scanned using a Thermo iBright imaging system. Assessment of CM/VLDL Contamination Blood samples were obtained under fasting conditions and 2 h after the meal from 6 healthy volunteers. The blood samples were centrifuged at 1300 × g for 10 min at 4°C, and the supernatant was transferred to a clean tube and further centrifuged at 16 000 × g for 10 min at 4°C. A sample of blood plasma 30 min after high‐fat meal was used for the purification of CM/VLDL, as was shown in the previous study [27]. 3 mL of blood plasma was centrifuged at 436 000 × g for 30 min at room temperature to remove cells and debris. The supernatant was transferred to a new tube without disturbing the pellet. Then, the blood plasma was centrifuged at 436 000 × g for 30 min at room temperature again to remove cells and debris. The supernatant was discarded, and the precipitated CM/VLDL was resuspended in 900 μL of PBS. Purified CM/VLDL solution was diluted 30‐fold with plasma for LD measurement. Nanoparticle Tracking Analysis NanoSight LM10 (Malvern, UK) was used for NTA measurement by following the manufacturer's protocol, consisting of a conventional optical microscope, Marlin charged coupled device camera, and a LM10 unit (sample unit) with a laser light source. Briefly, the motion of nanoscale particles (20–1000 nm) is illuminated by combining the properties of both laser light scattering microscopy and Brownian motion. The trace of the particles is recorded by a camera that captures the scattered light, and the Stokes‐Einstein equation is applied to determine the size distribution of each tracked particle, which enables the calculation of their size independent of density. In this study, a sample was diluted with deionized and filtrated water up to 3 mL to reach an optimal particle concentration of 107–109 particles/ml. The samples were injected into the sample unit (approximately 300 μL) with a 1 mL sterile syringe. The camera level, defined as the sensitivity and length of time the camera shutter is open, was increased until all particles were distinctly visible and did not exceed a particle signal saturation of 20% (manually set to level 10). The NanoSight Software NTA (version 3.00068) was used for data capturing and data evaluation. The NanoSight LM10 captured and recorded 60‐s videos, and the recordings were repeated until 5 videos had been captured. Laser Diffraction Particle size distribution was measured using an Aggregates Sizer SALD‐7500nano (Shimadzu Corp., Kyoto, Japan), following the manufacturer's protocol. This system utilizes a violet semiconductor laser with a wavelength of 405 nm, enabling high‐resolution measurement of particles ranging from 7 nm to 800 μm. A group of particles is irradiated by the laser, and the resulting diffraction and scattering patterns are detected. The system continuously calculates particle size distributions based on a quantitative optical model using Mie scattering theory [28]. It assumes that multiple scattering is negligible at the concentrations used, and the total scattered light is considered to be the sum of scattering from individual particles. The measurement duration can be as short as 1 s, and particle size distributions are displayed in real‐time at 1‐s intervals, using a single light source and wide‐range detection method. In our protocol, 65 μL of sample was mixed with 65 μL of PBS and placed in a microcell to ensure that the background absorbance remained below 0.2. Measurements were performed using WingSALD bio‐SCM software (version 3.4.7). The refractive index was set to 1.40 for both liposomes and EVs, and the cut‐off level was set to 100. The particle size was expressed as the median value of the measurement bins. Importantly, particle size distribution was evaluated under the assumption of a log‐normal distribution, which is the default statistical model applied by the WingSALD II software (Shimadzu Corp., Kyoto, Japan). This assumption was used for post‐processing and reporting the particle size characteristics shown in Figures 3 and 5. Statistical Analysis Analyses were performed using the R software program (version 3.6.1; The R Foundation for Statistical Computing, Vienna, Austria; http://www.R‐project.org), Microsoft Excel 2013, and Graphpad Prism 8. Data were summarized as mean ± standard deviation. For the linearity evaluation of the dilution series, a single regression line was fitted, and the regression line equation and coefficient of determination (R2) were calculated. For reproducibility evaluation, the coefficient of variation (CV) was calculated. For comparison between the two groups, a two‐tailed t‐test (Welch's t‐test) with unequal variances was used. All p‐values were two‐sided, and p < 0.05 was considered statistically significant. Receiver operating characteristic (ROC) curve analysis was performed to assess the discriminatory performance of plasma EV‐sized particle concentration for identifying patients with AF. The optimal cut‐off value was determined using the Youden index. Results Measurement of Synthetic Liposomes The size distribution and the number of particles were measured by LD and NTA using a synthetic liposome with a mean diameter of 100 nm. LD showed a slightly larger particle size distribution than NTA. The mean diameter of the particle was 106.8 nm in LD and 105.8 nm in NTA, and the mode diameter was 89.1 nm in LD and 103.6 nm in NTA (Figure 1a). The same liposome sample was measured 3 times by LD and NTA (Table 1). The coefficient of variation of LD measurement was 0.013% in the mean particle size and 2.9% in the number of particles, which indicated that the measurement by LD showed high reproducibility. Then the liposome samples were serially diluted and measured by LD and NTA (Figure 1b). Since the optimal measurement range of concentration was different between LD and NTA, we measured the liposome size using LD in 10 times higher concentration of liposome solution than that of NTA. Both LD and NTA showed high linearity for the measurement of dilution series (R2 = 0.9978 and R2 = 0.9635, respectively). Measurement of synthetic liposomes by NTA and LD. (a) Representative particle size distribution of synthetic liposomes (100 nm) measured by NTA and LD. Horizontal axis represents particle size (nm), and vertical axis represents particle concentration measured by NTA (1 × 106 particles/mL), and by LD (1 × 109 particles/mL). LD shows a similar average value of particle size as NTA. (b) Concentration of synthetic liposomes measured by NTA and LD in a dilution series of liposomes. Linearity of concentration was assessed by coefficient of determination. Dilution factors were different between measurements by NTA and LD. Data are shown as mean ± SD (n = 3). Error bars indicate SD; SD was smaller than the symbol size for the NTA measurement at a dilution of 1/20000. LD: Laser diffraction, NTA, nanoparticle tracking analysis; PBS, phosphate‐buffered saline. TABLE 1Reproducibility evaluation of particle size measurement by NTA and LD.NTALDDiameterSD (nm)4.830.014CV (%)4.420.0134BIAS (%)2.444.74NumberSD (1 × 109/ml)0.170.12CV (%)1.642.98Abbreviations: CV: coefficient of variation, LD: laser diffraction, NTA: nanoparticle tracking analysis, SD: standard deviation. Reproducibility evaluation of particle size measurement by NTA and LD. Abbreviations: CV: coefficient of variation, LD: laser diffraction, NTA: nanoparticle tracking analysis, SD: standard deviation. Measurement of EVs Extracted From Human Blood Plasma Next, we performed the measurement using EV fractions extracted and purified from human blood. Following the Minimal Information for Studies of Extracellular Vesicles (MISEV 2023) [29], we combined transmission electron microscopy (TEM), Western blotting, and two single‐particle techniques—LD and NTA. TEM confirmed the presence of the expected round‐to‐cup‐shaped vesicles, 50–150 nm in diameter, with no visible protein aggregates or cellular debris (Figure 2a). To satisfy the molecular criteria, Western blotting revealed strong bands for the surface tetraspanin CD63 and the cytosolic ESCRT‐associated protein Alix (Figure 2b). Detecting a transmembrane marker together with an internal marker fulfills the requirement for at least one protein from each topological category. Original, uncropped images of the Western blot membranes and gels related to Figure 2b are available in Figure S1. Measurement of EVs extracted from blood plasma by NTA and LD. (a) TEM image of EVs isolated from plasma. Vesicles display the characteristic round‐ to cup‐shaped morphology and fall within expected diameter range for small EVs (50–150 nm). Scale bar: 100 nm. (b) Western blot analysis of EV‐specific protein markers. EVs isolated from plasma were lysed and subjected to Western blotting. Specific bands for the tetraspanin CD63 and the cytosolic ESCRT‐associated protein Alix, well‐established markers for EVs, were detected, confirming successful isolation of EVs. (c) Representative particle size distribution of EVs extracted from blood plasma by NTA and LD. (d) Concentration of extracted EVs from plasma by NTA and LD in a dilution series. Linearity of concentration was compared by using the coefficient of determination (R2). Both NTA and LD showed high linearity. Data are shown as mean ± SD (n = 3). Error bars indicate SD; SD was smaller than the symbol size for LD at 1/2 and 1/4 dilutions and for NTA at 1/20 and 1/40 dilutions. [e] Averaged particle size of extracted EVs in a dilution series. Particle size tends to be larger in higher dilution rates both in measurements by NTA and LD. Data are shown as mean ± SD (n = 3). Error bars indicate SD; for LD measurements at dilutions of 1/2, 1/4, 1/8, and 1/16, and for NTA measurements at a dilution of 1/40, SD was smaller than the symbol size and thus not clearly visible. EVs, extracellular vesicles; TEM, transmission electron microscopy. Other abbreviations are the same as Figure 1. The clear identification of vesicle‐like structures also provides complementary validation of our particle size measurements evaluated by LD and NTA. LD showed a unimodal distribution pattern, similar to NTA, but the diameter of particle tended to be smaller in LD (mean diameter; 110 nm in LD and 149 nm in NTA, mode diameter: 101 nm in LD and 136. nm in NTA) (Figure 2c). Measurement of serially diluted samples showed that both LD and NTA had high linearity for the concentration of particles (R2 = 0.9925 in LD and R2 = 0.9910 in NTA) (Figure 2d). The mean particle size in serially diluted samples showed that both LD and NTA tended to show larger particle size with higher dilution (Figure 2e). The result of LD in measuring the EVs extracted from blood plasma indicated that LD had reproducibility comparable to NTA. Measurement of EV‐Sized Particles in Human Blood Plasma Without an Extraction Process Because EV extraction is cumbersome and may reduce quantitative accuracy, we examined whether LD could be applied to the measurement of EV‐sized particles in plasma without an extraction process. The particle size distribution was similar to that of EV fractions extracted from plasma and showed a peak at 101 nm (Figure 3a). Concentration measurements of serially diluted plasma samples showed high linearity, as observed for extracted EV fractions (R2 = 0.9519) (Figure 3b). The mean particle diameter of diluted plasma samples tended to increase with greater dilution, similar to the findings in extracted EV fractions (Figure 3c). We then compared measurements before and after extraction. Peripheral blood was collected from 10 healthy volunteers under fasting conditions. Plasma samples were prepared by conventional centrifugation, and EV fractions were prepared from the same plasma using the precipitation method. The particle concentration at the mode diameter, representing the most frequently detected size by LD, was 2.57 × 1010 particles/mL in plasma samples and 1.84 × 1010 particles/mL in extracted EV fractions. This decrease in peak particle concentration suggests partial particle loss during the extraction process (Figure 3d). Because the input plasma volume and the final resuspended volume after precipitation were identical, this difference was likely attributable to loss during extraction. However, the particle size distribution was similar before and after extraction, as shown by the normalized profiles in Figure 3e. Measurement of EV‐sized particles in blood plasma by LD without an extraction process. (a) Representative particle size distribution of EV‐sized particles in blood plasma measured by LD. Trace shows a unimodal pattern. (b) Concentration of EV‐sized particles in blood plasma measured by LD in a dilution series of plasma. Linearity of concentration was compared by using the coefficient of determination (R2). Data are shown as mean ± SD (n = 3). Error bars indicate SD; for LD measurements at dilutions of 1/8 and 1/32, SD was smaller than the symbol size and thus not clearly visible. (c) Averaged particle size measured by LD in a dilution series of plasma. Particle size tends to be larger at higher dilution rates. Data are shown as mean ± SD (n = 3). Error bars indicate SD; for LD measurements at dilutions of 1/2, 1/8, 1/16, and 1/32, SD was smaller than the symbol size and thus not clearly visible. (d) Particle size distribution of EV‐sized particles in blood plasma before and after extraction. Blood samples were collected from 10 healthy volunteers under fasting conditions. EV‐sized particles in plasma were measured by LD before extraction. EV fractions were then extracted from plasma using the precipitation method and measured by LD after extraction. (e) Normalized particle size distribution of EV‐sized particles measured before and after the extraction. After normalization at the peak value, the two profiles were almost identical. Abbreviations are the same as Figure 2. Effect of the Contamination of Chylomicrons and Very‐Low‐Density Lipoproteins on the Measurement of Blood Plasma Because some lipoproteins have size ranges overlapping with those of EVs, they may interfere with LD‐based measurement in plasma samples without an extraction process [30]. To assess this effect, blood samples were obtained at the fasting state and 2 h after ingestion of a high‐fat diet from 6 healthy volunteers. The distribution of particle size in the fasting samples showed a peak at 101 nm, consistent with previous results. However, samples obtained 2 h after ingestion of a high‐fat diet showed a shift in the particle size distribution toward larger sizes, with a peak at 189 nm (Figure 4a). Effect of CM/VLDL contamination on LD‐based measurement in plasma without an extraction process. (a) Particle size distribution of EV‐sized particles in plasma measured by LD before and after meal intake. Blood samples were collected in the fasting state and 2 h after ingestion of a high‐fat diet. (b) Representative particle size distribution of purified CM/VLDL sample. Trace shows a unimodal pattern with a peak at 189 nm. (c) Representative particle size distributions measured by LD in fasting blood plasma with or without purified CM/VLDL sample. Fasting plasma mixed with purified CM/VLDL fraction shows a positive shift of peak. (d) Normalized particle size distribution of EV‐sized particles in plasma before and 2 h after ingestion of high‐fat diet (the same samples as Figure 4a), and fasting plasma mixed with purified CM/VLDL (the same sample as Figure 4c). Distribution curves are normalized to their peak values. Particle size distribution curve of 2 h after meal is identical to that of fasting plasma mixed with purified CM/VLDL. CM, chylomicron; VLDL, Very‐low‐density lipoprotein. Then we measured the particle size and concentration of chylomicrons and very‐low‐density lipoproteins (CM/VLDL) solution by LD using samples purified from blood plasma 30 min after the intake of a high‐fat meal. The particle size distribution of purified CM/VLDL showed a unimodal pattern with a peak at 189 nm (Figure 4b). Then, we mixed the CM/VLDL solution with plasma obtained at the fasting state and measured it by LD. The distribution of particle size shifted to a larger size, with its peak at 189 nm (Figure 4c). This distribution profile was normalized and overlapped with the plasma obtained after the meal, which showed an identical pattern (Figure 4d). These results indicated that postprandial blood samples affected LD‐based measurement in plasma without an extraction process because of contamination by CM/VLDL. EV‐Sized Particle Concentration in Fasting Plasma in Patients With Atrial Fibrillation LD‐based measurement in plasma without an extraction process was then applied to fasting blood samples from 20 patients with AF and 20 controls without a history of AF. Baseline characteristics are summarized in Table 2. At the time of blood sampling, 7 of 20 patients with AF (35%) were in AF rhythm, whereas all control patients were in sinus rhythm. Age, sex, and BMI were comparable between the groups, while BNP was significantly higher in patients with AF (Table 2). The particle size distribution profile was similar between AF patients and controls (Figure 5a). However, the concentration of EV‐sized particles was significantly higher in patients with AF than in controls (Figure 5b). ROC curve analysis of plasma EV‐sized particle concentration showed modest discriminatory performance for identifying patients with AF, with an area under the curve of 0.626 (95% confidence interval, 0.445–0.807). The optimal cut‐off value was 2.21 × 1011 particles/mL, yielding a sensitivity of 42.1% and a specificity of 84.2% (Figure.S2). TABLE 2Baseline characteristics of controls without AF and patients with AF.CharacteristicsControls (N = 20)AF (N = 20) p valueAge, years73 [62–78]72 [64–76]0.704Male, n (%)12 (60%)13 (65%)0.98BMI, kg/m2 23.1 [21.2–25.2]22.4 [20.1–23.7]0.279Sinus rhythm at sampling, n(%)20 (100%)13 (65%)—AF at sampling, n (%)0 (0%)7 (35%)—Paroxysmal AF—17 (85%)—Persistent or long‐standing AF—3 (15%)—ComorbiditiesHypertension, n (%)14 (70%)5 (25%)0.01Diabetes mellitus, n (%)8 (40%)0 (0%)0.003Dyslipidemia, n (%)16 (80%)3 (15%)< 0.001Coronary artery disease, n (%)12 (60%)0 (0%)< 0.001Heart failure, n (%)0 (0%)7 (35%)0.008Chronic kidney disease, n (%)5 (25%)5 (25%)0.98Prior stroke/TIA, n (%)1 (5%)1 (5%)0.98Smoking, n (%)3 (15%)6 (30%)0.451Laboratory dataTriglycerides, mg/dL102 [76–156]96 [73–176]0.922LDL cholesterol, mg/dL91 [72–116]112 [94–126]0.089HDL cholesterol, mg/dL68 [58–82]74 [53–82]0.955eGFR, mL/min/1.73 m2 63.9 [50.1–71.6]62.2 [52.7–68.4]0.372CRP, mg/dL0.06 [0.03–0.12]0.04 [0.03–0.08]0.268BNP, pg/mL36.9 [18.4–47.6]150.5 [33.8–297.4]0.014MedicationOral anticoagulants, n (%)1 (5%)20 (100%)< 0.001Statins, n (%)10 (50%)3 (15%)0.041Antiarrhythmic drugs, n (%)0 (0%)10 (50%)< 0.001 Note: Continuous variables are presented as median [interquartile range, 25th–75th percentile], and categorical variables as n (%). Blood samples were collected on the day of admission after an overnight fast of at least 8 h. Sampling rhythm is presented as sinus rhythm or AF. Among the 20 patients with AF, 17 had paroxysmal AF and 3 had persistent or long‐standing persistent AF; the duration of persistent AF in these 3 patients was 6 months, > 10 years, and 2 years, respectively.Abbreviations: AF, atrial fibrillation; BMI, body mass index; BNP, B‐type natriuretic peptide; CRP, C‐reactive protein; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; TIA, transient ischemic attack. Baseline characteristics of controls without AF and patients with AF. Note: Continuous variables are presented as median [interquartile range, 25th–75th percentile], and categorical variables as n (%). Blood samples were collected on the day of admission after an overnight fast of at least 8 h. Sampling rhythm is presented as sinus rhythm or AF. Among the 20 patients with AF, 17 had paroxysmal AF and 3 had persistent or long‐standing persistent AF; the duration of persistent AF in these 3 patients was 6 months, > 10 years, and 2 years, respectively. Abbreviations: AF, atrial fibrillation; BMI, body mass index; BNP, B‐type natriuretic peptide; CRP, C‐reactive protein; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; TIA, transient ischemic attack. LD‐based measurement of EV‐sized particles in blood plasma in patients with atrial fibrillation. (a) LD‐based measurement of EV‐sized particles in fasting plasma samples from patients with AF and controls without AF. Both groups showed a similar unimodal pattern, whereas the particle concentration was higher in the AF group. (b) Concentration of EV‐sized particles in fasting plasma from patients with AF and controls without AF. The AF group showed a significantly higher particle concentration than the control group. AF, atrial fibrillation; EV, extracellular vesicles. Discussion To the best of our knowledge, this is the first study to demonstrate the validity of quantitative measurement of EV‐sized particles using LD and to establish a simple method for plasma measurement without an extraction process. Although LD has traditionally been applied to characterize particle size distributions in various colloidal systems, recent studies have shown that LD can reliably assess complex multimodal particle populations [31, 32, 33]. In this study, we evaluated particle size and concentration using synthetic liposomes and EVs extracted from blood plasma. The accuracy of the measurement of LD was evaluated by comparison with NTA, a commonly used method for EV measurement. Although the optimal concentration range for the measurement was different, we found that the size and concentration measurement by LD was comparable with that by NTA, and LD had high reproducibility in repeated measurements. However, LD has the disadvantage that the resolution for particle size was lower than NTA. In addition, the measurement of particle size changed depending on the number of particles, both in LD and NTA. Thus, care must be taken when comparing samples of different concentrations. We then performed measurement of EV‐sized particles in peripheral blood plasma without an extraction process. Since LD showed an identical particle size distribution profile between plasma samples measured with and without extraction, this method appeared applicable to plasma measurement without an extraction process, consistent with recent reports showing the utility of LD for detecting particle agglomeration or protein contamination [33]. The particle concentration was higher in plasma samples measured without extraction than in EV fractions extracted and resuspended from plasma. This difference may reflect partial particle loss during the precipitation and centrifugation steps. Therefore, measurement in plasma without an extraction process may improve quantitative accuracy by avoiding sample loss during preprocessing, while also reducing the effort and cost associated with EV extraction. Since EVs share several physicochemical properties with lipoproteins [34, 35, 36], contamination of lipoproteins can affect measurement results. In particular, given the reported particle size ranges of approximately 80–1000 nm in CM and 30–80 nm in VLDL, these particles may interfere with LD‐based measurement of EV‐sized particles. The measurement of fasting plasma reduced the contamination of CM/VLDL. We found that the contamination of CM/VLDL decreased the concentration of smaller particles. The LD measurement irradiated the laser to the group of particles in principle, and the existence of too much contamination of larger size particles concealed the fraction of EVs, resulting in an underestimation. This finding highlights the importance of fasting blood sampling to minimize CM contamination in LD‐based plasma measurement. However, because postprandial lipoproteins can substantially affect the measurement results, the need for fasting blood sampling may limit the feasibility of this approach in some routine clinical settings. We applied LD‐based measurement in fasting plasma samples from controls without AF and patients with AF. Of note, AF patients showed a significantly higher concentration of EV‐sized particles than controls. This finding is consistent with previous reports suggesting that circulating EV‐associated molecules, including miRNAs, may reflect AF‐related pathophysiology and have potential as circulating biomarkers [7, 8, 37]. From a clinical perspective, AF is frequently asymptomatic or paroxysmal, which complicates timely diagnosis and risk stratification [11]. From a broader perspective, there is growing interest in identifying clinically useful biomarkers grounded in the pathophysiology of arrhythmias [14]. Previous studies have explored inflammatory and structural biomarkers associated with AF onset and progression; however, their clinical applicability has been limited by variability and lack of standardization [13, 38]. A simple and reproducible method for EV quantification, such as LD‐based measurement of plasma, may therefore support future efforts to detect AF‐related circulating biomarkers and complement existing diagnostic approaches [39]. Although the present study was not designed to address causal mechanisms, further investigations may clarify how alterations in EV‐sized particle concentration relate to atrial remodeling and disease progression in AF [40]. Although ROC analysis showed only modest discriminatory performance for identifying patients with AF, elevated EV‐sized particle levels may still have clinical relevance beyond diagnosis itself. A recent study showed that circulating EVs released in AF are taken up by endothelial cells, leading to impaired endothelial function. Accordingly, elevated EV levels may not serve as a diagnostic marker of AF, but rather as a predictor of the risk of associated complications [41]. Several limitations of this study should be acknowledged. First, due to physical constraints inherent to LD, accurate detection of particles smaller than approximately 200 nm remains challenging. Although the SALD‐7500nano system theoretically estimates particles down to ~100 nm using Mie theory‐based fitting, measurements of smaller extracellular vesicles should be interpreted with caution. Nevertheless, LD‐derived size distributions were generally consistent with NTA and supported by TEM. Second, synthetic liposomes were used as surrogate particles to evaluate LD performance. Because liposomes have a more homogeneous internal composition and refractive index than native EVs, differences in optical properties may influence scattering behavior and particle size estimation. Third, LD‐based measurement in plasma without an extraction process may be affected by particles with overlapping size ranges, particularly lipoproteins such as CM/VLDL. Although fasting blood sampling minimized this effect in the present study, potential interference from other circulating particles cannot be completely excluded. Fourth, the sample size was small, and circulating particle concentration in plasma may be influenced not only by AF itself but also by comorbidities and differences in clinical background between the groups. In addition, LD is not specific for EVs and may detect other circulating particles with overlapping size ranges and similar optical properties. Consistent with these limitations, ROC analysis showed only modest discriminatory performance of plasma particle concentration for identifying patients with AF. Therefore, the present findings should be interpreted as exploratory and hypothesis‐generating. Moreover, because LD does not provide information on the cellular origin or molecular cargo of the detected particles, the biological significance of the increased EV‐sized particle concentration in AF remains unclear. Conclusions In conclusion, LD enables highly reproducible quantification of extracellular vesicle‐sized particles in blood plasma without an extraction process. Measurement using fasting plasma may provide a practical proof‐of‐concept approach for biomarker assessment in atrial fibrillation. Author Contributions G.N., K.T., and T.S. led the research concept and design, conducted data collection and analysis, and drafted the manuscript. N.U, S.I., S.H., A.S., R.I., K.K., R.K, J.Y, K.Y., T.N., and T.Y. supported data analysis, provided critical advice on the methodology, and reviewed the manuscript. K.I. and R.O. conducted experiments and data collection, contributed to the interpretation of results, and assisted in manuscript editing. T.S. managed the project, and supervised the overall study. Funding This study was supported in part by the Grant‐in‐Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan (No. 22K08152 to T.S.). Conflicts of Interest The authors declare no conflicts of interest. Supporting information Figure S1: Original uncropped Western blot images. Uncropped images of the membranes and gels used for Western blot analysis in Figure 2b are shown. The full membrane images for CD63 (FigureS1a) and Alix (FigureS1b) are provided to ensure transparency and demonstrate the integrity of the experimental results. Figure S2: Receiver operating characteristic curve of plasma EV‐sized particle concentration for identifying patients with AF. The area under the curve was 0.626 (95% confidence interval, 0.445–0.807). The optimal cut‐off value determined by the Youden index was 2.21 × 1011 particles/mL, yielding a sensitivity of 42.1% and a specificity of 84.2%.