Early Diagnosis of Pancreatic Ductal Adenocarcinoma by Signal-Enhanced Lateral Flow Immunoassay: SELFI

Early Diagnosis of Pancreatic Ductal Adenocarcinoma by Signal-Enhanced Lateral Flow Immunoassay: SELFI | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Early Diagnosis of Pancreatic Ductal Adenocarcinoma by Signal-Enhanced Lateral Flow Immunoassay: SELFI Bong-Hyun Jun, Sohyeon Jang, Minsup Shin, Jiseok Han, Han-Joo Bae, and 11 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6695327/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 05 Feb, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Abstract Pancreatic ductal adenocarcinoma (PDAC) is linked to high incidence and mortality rates because it is often detected in later stages, when the prognosis is poor. However, the current state-of-the-art methods for diagnosing early PDAC tend to be invasive, time-consuming, and unreliable, primarily due to the difficulties associated with the early detection of pancreatic cancers. Here, we report a quick and sensitive method for the early diagnosis of PDAC using a signal-enhanced lateral flow immunoassay called SELFI. We developed SELFI, which can generate a strong colorimetric signal through multiple hotspots formed by plasmonic gold nanoparticles (AuNPs) assembled on a silica bead. Our SELFI assay achieved a 28-fold increase in the limit of detection compared to conventional lateral flow immunoassays using 20 nm AuNPs, providing results within 15 min. We demonstrated that SELFI can be utilized for the early diagnosis of PDAC, as indicated by a receiver operating characteristic curve and a larger area under the curve compared to the enzyme-linked immunosorbent assay (**** P < 0.0001). SELFI's effective diagnostic features could enhance the timely identification of PDAC and may also serve in the early diagnosis of a range of other diseases. Biological sciences/Cancer/Paediatric cancer Biological sciences/Biotechnology/Nanobiotechnology/Nanostructures Biological sciences/Biotechnology/Nanobiotechnology/Biosensors Figures Figure 1 Figure 2 Figure 3 Figure 4 INTRODUCTION Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest solid tumors, characterized by a persistently poor prognosis 1 , 2 . Over the past few decades, the incidence of PDAC has steadily increased; however, the 5-year survival rate has remained below 10% 1, 3 . This grim prognosis is primarily due to three factors: difficulties in early detection, the aggressive biology of the tumor, and resistance to current therapies 4 , 5 . Among various approaches to overcoming these obstacles, early detection and curative resection are the most critical and effective strategies 6 , 7 . Furthermore, even in postoperative patients, early detection of minimal residual disease or small recurrent tumors is essential for timely intervention 7 – 9 . Despite its importance, early diagnosis of PDAC is challenging because significant symptoms are not observed in the early stages, and the pancreas is difficult to examine using standard diagnostic methods due to its deep anatomical location 5 , 10 . PDAC is currently diagnosed using pancreatic imaging techniques such as computed tomography, magnetic resonance imaging, and endoscopic ultrasound (EUS) examinations 11 . If a pancreatic tumor is not identified using these methods, invasive diagnostic approaches, such as EUS-guided fine needle aspiration (EUS-FNA), may be required to confirm the presence of a tumor 12 . However, these complex and invasive procedures can be burdensome for patients, leading to reluctance and delays in diagnosis, resulting in missed optimal diagnostic opportunities. Therefore, alternative diagnostic methods that are more convenient and accessible compared to existing techniques are essential for the early diagnosis of PDAC. As a noninvasive approach, diagnosing pancreatic ductal adenocarcinoma (PDAC) using biomarkers in biological samples such as serum has been proposed. Numerous biomarkers for PDAC detection have been investigated; however, cancer antigen 19 − 9 (CA19-9) remains the only FDA-approved biomarker 13 – 15 (Fig. 1 a). CA19-9 has demonstrated reliability in evaluating treatment responses in advanced PDAC; however, it has limitations as a screening marker for early-stage resectable PDAC when used with existing biomarker screening systems, such as enzyme-linked immunosorbent assay (ELISA) 15 , 16 . To overcome the limitations of CA19-9, research efforts like CancerSEEK have aimed to discover new blood biomarkers beyond CA19-9 17, 18 . Despite these initiatives, the discovery of novel PDAC biomarkers presents challenges related to reproducibility and commercial viability. As an alternative, the development of a novel quantification system for CA19-9 to replace ELISA has been explored to enhance the sensitivity and efficacy of early PDAC diagnosis. One candidate method for CA19-9 quantification in PDAC diagnosis is the lateral flow immunoassay (LFIA), which is widely used for point-of-care testing due to its convenience. Generally, the LFIA system produces red-colored signals from gold nanoparticles (AuNPs) used as probes. The LFIA system is more user-friendly than conventional liquid biopsy methods, and the time necessary to obtain analysis results is very short 19 . However, the colorimetric signal intensity of the AuNP probe is insufficient, necessitating the development of a new probe 20 . Consequently, recent reports have shown that not only silver nanoparticles (NPs) but also alloy-type metal NPs—such as gold/platinum and gold/iridium—and even NPs with assembled nanostructures, in which quantum dots or metal NPs are assembled onto silica NPs, have been utilized as probes for LFIA 21 – 27 . Due to its great convenience, LFIA has been successfully applied to analyze various targets, such as exosomes 23 , prostate cancer antigen 24 , 25 , hormones 28 , carcinoembryonic antigen 29 , and the COVID-19 virus 26 . However, the quantitative analysis of CA19-9 using LFIA for early PDAC diagnosis has rarely been reported, and in the few existing studies, the limit of detection (LOD), which reflects the sensitivity of the analytical system, was not adequate for highly sensitive analysis 30 , 31 . In this study, we designed a Signal-Enhanced Lateral Flow Immunoassay (SELFI) system for the early diagnosis of PDAC through the quantitative analysis of CA19-9 in serum samples. Silica nanoparticles (SiO 2 NPs) with assembled gold nanoparticles (AuNPs) (SiO 2 @Au@Au NPs), which feature numerous hotspots due to nanogaps between the assembled AuNPs, were fabricated as the colorimetric nanoprobes in the SELFI system (Fig. 1 b and c). Our SELFI system required only 15 minutes after sample loading for the quantitative analysis of CA19-9 in serum samples, while the commonly used ELISA for biomarker quantification requires over 4 hours and involves a complex experimental process (Fig. 1 d). The colorimetric signal intensity of SELFI was significantly higher than that of conventional lateral flow immunoassays (LFIA) at the same CA19-9 concentration, thanks to the hotspots in SiO 2 @Au@Au NPs, which markedly enhanced colorimetric signal intensity (Fig. 1 e). The limit of detection (LOD) of CA19-9 in SELFI was 27.6-fold lower than that of conventional LFIA (Fig. 1 f). As a diagnostic system for PDAC based on the quantitative analysis of CA19-9, SELFI exhibited superior diagnostic performance compared to both ELISA and conventional LFIA, as indicated by their receiver operating characteristic (ROC) curve and area under the curve (AUC) values. Notably, SELFI demonstrated significant improvement in the early diagnosis of PDAC (Fig. 1 g). RESULTS Design of Colorimetric Nanoprobe for SELFI Using Numerical Simulation Based on previous studies, hotspots form between metal NPs, such as AuNPs or Ag/Au NPs, when they assemble on the surface of SiO 2 NPs (Fig. 2 a) 32 , 33 . However, the relationship between the intensity of the electric field—enhanced by the hotspot effect—and parameters like the gap distance between the AuNPs ( G Au ) and the diameter of the AuNPs ( D Au ) has yet to be explored. Numerical simulations were performed to investigate the relationships between the optical properties of AuNPs-assembled silica NPs and these parameters, particularly G Au and D Au , to aid in the in silico design of innovative colorimetric nanoprobes for SELFI. Supplementary Fig. 1 illustrates a schematic of the 3D model geometry and the various cases considered in the simulation. The assembled nanostructure comprised SiO 2 NPs with a diameter of 168 nm, and 270 AuNPs were evenly distributed on its surface. The electric field was oriented in the y-direction. Six simulation cases were analyzed, with G Au values of 10, 8, 6, 4, 2, and 1 nm. Since the diameter of the SiO 2 NP and the count of AuNPs were fixed, D Au increased as G Au decreased. Using these designed assembled nanostructures, a wave-optics simulation was performed in the visible wavelength range, assuming that the assembled nanostructures were immersed in water. Based on the simulation results, the electric field norm spectra of each model in the visible wavelength range were analyzed (Supplementary Fig. 2a). The electric field on the surface of the assembled nanostructure became stronger as G Au decreased, and hotspots emerged between the AuNPs when G Au was less than 6 nm (Fig. 2 b). The results indicated that the overall intensity of the electric field increased with decreasing G Au , with a particularly notable increase at the peak wavelength. The extinction cross-sectional spectra, which represent the sum of the absorption and scattering cross-sections, also increased as G Au decreased and exhibited a peak at a wavelength similar to that of the electric field norm spectra (Fig. 2 c, Supplementary Fig. 2b, c). Although the extinction cross-section increased as G Au decreased, it remained unclear whether this increase was attributed to the reduction in G Au or the rise in D Au , as both parameters changed simultaneously. To determine which parameter had a more significant impact on increasing the extinction cross-section of the nanostructure, the change in the extinction cross-section of a single AuNP on the assembled nanostructure was compared with that of a single AuNP on a SiO 2 substrate (Supplementary Figs. 3 and 4, Fig. 2 d, e). As shown in Supplementary Fig. 4, the electric field of a single AuNP on the SiO 2 substrate increased as D Au increased. In contrast, the electric field from the AuNPs on the assembled nanostructure increased dramatically with the creation of hotspots between AuNPs when G Au was reduced to a certain distance (Fig. 2 d). Due to this phenomenon, the extinction cross-section of AuNPs in the assembled nanostructure was nearly seven times higher than that of a single AuNP on the SiO 2 substrate at the largest D Au , despite being nearly identical at the smallest D Au (Fig. 2 e). This comparison suggests that G Au , which is closely linked to the generation of hotspots between AuNPs, plays a more significant role in enhancing the electric field than D Au . We confirmed that the assembled nanostructure exhibited a significantly higher extinction intensity than a single 20 nm AuNP, the gold standard nanoprobe for conventional LFIA (Supplementary Fig. 5). These simulation results suggest that the amplification of the electric field around the assembled nanostructure, resulting in a strong extinction intensity, is not solely due to the increase in D Au but is mainly driven by the formation of hotspots caused by the decrease in G Au as D Au increases. Based on these simulation results, we could expect that our designed SiO 2 @Au@Au NPs should exhibit a more intense colorimetric signal due to the electric field enhancement effect caused by hotspots formed between AuNPs with under the specific distance. This enhancement renders them superior to single AuNPs, which are commonly used as nanoprobes in conventional LFIA system. Fabrication and Characterization of Nanoprobes Based on the simulation results, we fabricated SiO 2 @Au@Au NPs via a seed-mediated growth method, with controlling the nanogaps as colorimetric nanoprobes for SELFI (Fig. 2 f and Supplementary Fig. 6). AuNPs were grown after introducing 3–5 nm AuNPs onto the surface of aminated SiO 2 NPs, with control over the amount of Au precursor and ascorbic acid. Each particle was named as SiO 2 @Au@Au 1 to SiO 2 @Au@Au 11 according to the reagents used. Figure 2 g shows the Cs-corrected scanning transmission electron microscopy (TEM) (Cs-STEM) images of SiO 2 @Au@Au 1 , SiO 2 @Au@Au 3 , SiO 2 @Au@Au 5 , SiO 2 @Au@Au 7 , SiO 2 @Au@Au 9 , and SiO 2 @Au@Au 11 . The average G Au for each nanostructure was 9.55, 7.63, 5.89, 3.53, 1.33, and 1.20 nm, respectively. This decrease in G Au resulted from the increase in D Au , which was based on the amount of Au 3+ precursor used during the fabrication process (Supplementary Fig. 8). After fabricating SiO 2 @Au@Au NPs, the optical properties of each nanoprobe were characterized to select the optimal colorimetric nanoprobes for SELFI. First, the degree of visible light extinction, strongly related to the colorimetric signal intensity of the SELFI system, was measured. As shown in Fig. 2 h, the extinction of visible light by the SiO 2 @Au@Au NPs increased as G Au of SiO 2 @Au@Au NPs was decreased, following the same trend observed in the numerical simulation. Based on these extinction spectra, the color of the SiO 2 @Au@Au NP mixture deepened as G Au of SiO 2 @Au@Au NPs was decreased (Supplementary Fig. 9). Similarly, the intensity of the colorimetric signals from the SiO 2 @Au@Au NPs on the NC membrane exhibited the same trend, with the colorimetric signal intensity from SiO 2 @Au@Au 11 NPs nearly reaching the maximum measurable value (Fig. 2 i). To identify the experimental primary factor behind this increase, the extinction spectra of Au nanoseeds (3–5 nm), SiO 2 @Au@Au 11 NPs, and AuNPs obtained from SiO 2 @Au@Au 11 NPs after removing the SiO 2 nanotemplate were measured (Fig. 2 j). The effect of AuNP diameter was evaluated by comparing the extinction of Au nanoseeds with that of single AuNPs from SiO 2 @Au@Au 11 NPs, which could be regarded as individual AuNPs with large interparticle distances. The effect of gap distance was assessed by comparing SiO 2 @Au@Au 11 NPs with AuNPs from SiO 2 @Au@Au 11 NPs, as the AuNP diameter remained unchanged after SiO 2 nanotemplate removal. A comparison of the maximum extinction of each nanostructure at their λ max.ext showed that the degree of extinction increased slightly from 0.031 (Au nanoseeds) to 0.120 (single AuNPs from SiO 2 @Au@Au 11 NPs) due to the increase in AuNP diameter. However, the extinction increased significantly to 0.426 for SiO 2 @Au@Au 11 NPs, attributed to the localized surface plasmon resonance effect from the nanogaps between the AuNPs (Fig. 2 k). These results experimentally confirm that the extinction of visible light by the fabricated nanostructure is more influenced by the gap distance between AuNPs than by AuNP diameter. To verify the superior optical properties of SiO 2 @Au@Au 11 NPs as colorimetric nanoprobes, they were compared with 20 nm AuNPs for conventional LFIA, and 200 nm AuNPs, which are similar in diameter to SiO 2 @Au@Au 11 NPs. In terms of visible light extinction, the maximum extinction value of SiO 2 @Au@Au 11 NPs was 58-fold higher than that of 200 nm AuNPs and 1,439-fold higher than that of 20 nm AuNPs (Supplementary Fig. 10). As shown in Supplementary Fig. 11, when each type of nanoprobe was observed using dark-field microscopy, only SiO 2 @Au@Au 11 NPs appeared brightly due to the surface plasmon resonance effect originating from the hotspots of the assembled nanostructure, whereas the others, which lacked hotspots, appeared faintly. Finally, the colorimetric signal intensity of SiO 2 @Au@Au 11 NPs on the NC membrane was significantly higher than that of 20 nm AuNPs on the NC membrane when the particle concentration exceeded a certain level (Supplementary Fig. 12). In summary, our fabricated SiO 2 @Au@Au 11 NPs exhibit superior optical properties as colorimetric nanoprobes compared to conventional AuNPs. Quantification of CA19-9 Using SELFI After fabricating and characterizing SiO 2 @Au@Au NPs, a model test was conducted for the quantitative analysis of CA19-9 using the SELFI system, which has the potential to be faster than ELISA and more sensitive than LFIA (Fig. 3 a). First, the change in colorimetric signal over time was examined to determine the optimal analysis time for SELFI. As shown in Fig. 3 b, the intensity of the test line's colorimetric signal increased with time, with no further significant changes observable by the naked eye after 15 minutes. This trend was further confirmed by quantifying the colorimetric signal intensity using ImageJ, which demonstrated that the intensity saturated after 15 minutes. These results indicate that quantification of CA19-9 in the sample using SELFI can be completed in just 15 min.—significantly shorter than the 285 minutes required for ELISA. Next, the colorimetric signals obtained from the SELFI, LFIA, and ELISA systems at various CA19-9 concentrations (0–200 U mL − 1 ) were compared (Fig. 3 c). In the SELFI system, represented by a navy-colored line as a signal, the colorimetric signal of test line signal was detectable by the naked eye when the CA19-9 concentration exceeded 5 U mL − 1 . However, no visible signals were detected in the LFIA system until the CA19-9 concentration reached 100 U mL − 1 , and even at this concentration, only a faint red line was observed. These tendencies were further confirmed by quantifying the colorimetric signal intensity using the ImageJ software. For the SELFI system, the signal intensity was 6.34 at a CA19-9 concentration of 1 U mL − 1 , increasing 14-fold to 90.2 at 200 U mL − 1 . This increase was significantly greater than that observed in the LFIA system, where the signal intensity increased only two-fold, from 8.81 to 19.15, over the same CA19-9 concentration range. Considering that ELISA, a well-established high-sensitivity detection method for target biomarkers, exhibited a similar signal increase pattern, these results suggest that the SELFI system offers a sensitivity comparable to that of ELISA for the detection of target biomarkers, including CA19-9. After verifying the fundamental characteristics of the SELFI system, calibration curves were obtained for the quantification of CA19-9 using each nanoprobe and corresponding standard solutions. The colorimetric signals from the test lines of both nanoprobes were measured using the ImageJ software, and the calibration curve of each analysis method was determined through sigmoidal fitting (Fig. 3 d and Supplementary Fig. 14). The equation for the calibration curve for SELFI was as follows: $$\:Signal\:intensity=83.84+\frac{(14.75-83.84)}{1+{\left(\frac{Concentration}{20.62}\right)}^{0.85}}$$ For single AuNP-based LFIA, the calibration curve was: $$\:Signal\:intensity=31.13+\frac{(6.45-31.13)}{1+{\left(\frac{Concentration}{689.82}\right)}^{1.26}}$$ The LOD of each system was calculated based on these calibration curves. The LOD for CA19-9 using SELFI was 0.15 U mL − 1 , which was 27.6-fold lower than that of single AuNP-based LFIA (4.14 U mL − 1 ) and two-fold lower than that of commercially available ELISA (0.3 U mL − 1 ). Furthermore, the LOD of SELFI was lower than that of previously reported LFIA systems for CA19-9 quantification 31 , indicating that SELFI is among the most sensitive LFIA systems for CA19-9 detection. Similar trends were observed not only for CA19-9 but also for prostate-specific antigen (PSA) and intercellular adhesion molecule 1 (ICAM-1), which are frequently used as cancer biomarkers. Therefore, the SELFI system could serve as a substitute for conventional LFIA systems, regardless of the target biomarker type (Fig. 3 e, f). To further assess the performance of SELFI as a diagnostic system, CA19-9 quantification-based PDAC diagnostic methods were compared by plotting analysis time versus LOD (Fig. 3 g). The results demonstrated that SELFI quantification of CA19-9 required a short analysis time, comparable to conventional LFIA 34 , LFIA with carbon nanotubes (1) 30 , and time-resolved fluorescence microsphere-based LFIA (2) 31 , all of which have relatively high LODs. Conversely, SELFI exhibited a lower LOD than or similar to that of commercially available ELISA kits, electrochemical immunosensors (3, 4, and 5) 35 – 37 , and near-infrared photothermal immunoassays (6) 38 , which require longer analysis time. These results highlight that SELFI effectively combines the advantages of different diagnostic systems, providing both rapid analysis and a low LOD. Diagnosis of PDAC via CA19-9 Quantification in Serum Samples Using SELFI Using LFIA and the designed SELFI system, quantification of CA19-9 in serum samples from healthy controls (50 ea.), patients in early-stage PDAC (40 ea.), and patients in late-stage PDAC (60 ea.) was conducted to determine the possibility of early diagnosis of PDAC using the SELFI system. After taking photographs of the test strips, the concentration of CA19-9 in each serum sample was successfully calculated by measuring the colorimetric signal intensity of the test lines using the ImageJ software (Supplementary Fig. 15, 16, and Supplementary Table S2). To compare the diagnostic performance of SELFI with that of ELISA, which is commonly used for the quantification of CA19-9, and LFIA, the ROC curve and AUC values of each system were compared. The ROC curve and AUC value were calculated after 90% of the samples were extracted from the mother populations, and this process was repeated 100 times to enhance statistical reliability (Fig. 4 a). Figure 4 b–d shows the average ROC curves of each system, in which the mother population was set as samples from healthy controls and early-stage patients (Fig. 4 b), healthy controls and late-stage patients (Fig. 4 c), and total samples (Fig. 4 d). The average AUC value of SELFI was significantly higher than that of ELISA for samples from healthy controls and early-stage patients (0.898 and 0.854, respectively), indicating that our SELFI system performed notably better for the early diagnosis of PDAC than ELISA. The AUC values of both diagnostic systems were almost the same for samples from healthy controls and late-stage patients (0.925 for SELFI and 0.926 for ELISA, respectively). For the total serum samples, the AUC value of SELFI was slightly higher than that of ELISA (0.913 for SELFI and 0.896 for ELISA, respectively). Finally, LFIA, which used 20 nm AuNPs as colorimetric nanoprobes, showed poor AUC values in all cases compared to the other two systems (0.762 for healthy controls and early-stage patients, 0.834 for healthy controls and late-stage patients, and 0.805 for total samples). The calculated AUC values for all repetition steps were plotted as violin plots to confirm the statistical evaluation of the differences in AUC values among ELISA, LFIA, and SELFI (Fig. 4 e–g). P values were less than 0.0001, except in the case of SELFI and ELISA for healthy controls and late-stage patients (0.3535), and the 95% confidence intervals (CIs) of the AUC values from SELFI were narrow (Table 1 ). Based on these results, the superior performance of the SELFI system for PDAC diagnosis, especially for early diagnosis, was statistically proven. Table 1 AUC calculations for ELISA, LFIA, and SELFI results at 95% CIs (100 repetitions) Samples Diagnosis system ELISA LFIA SELFI AUC 95% CI AUC 95% CI AUC 95% CI Early-stage (n = 40) + Healthy control (n = 50) 0.854 0.834–0.881 0.762 0.736–0.792 0.898 0.878–0.924 Late-stage (n = 60) + Healthy control (n = 50) 0.926 0.917–0.940 0.834 0.815–0.861 0.925 0.914–0.949 Early-stage (n = 40) + Late-stage (n = 60) + Healthy control (n = 50) 0.896 0.885–0.913 0.805 0.789–0.822 0.913 0.900–0.931 CONCLUSION We designed and fabricated SiO 2 @Au@Au NPs, where Au NPs were embedded onto the surface of SiO 2 NPs, as colorimetric nanoprobes for the LFIA system for the early diagnosis of PDAC. The SiO 2 @Au@Au NPs demonstrated more intense colorimetric signals than the commercially used single Au NPs, due to the plasmonic enhancement effect from the nanogaps between the Au NPs. Based on the most intense colorimetric signal observed at the closest gap distance, SiO 2 @Au@Au 11 NPs were selected as the optimal nanoprobes among the fabricated NPs. Similar to other LFIA systems, the quantification of CA19-9 was performed within 15 minutes, and LOD of SELFI was 0.15 U mL − 1 , which was 27.6-fold lower than that of single AuNPs. The concentration of CA19-9 in serum samples was determined after SELFI, and the AUC value of the ROC curve from SELFI was 0.913, which was higher than the values of 0.896 from ELISA and 0.805 from LFIA (**** P < 0.0001). Notably, the differences in AUC values between SELFI-ELISA and SELFI-LFIA became more pronounced when samples from healthy controls and early-stage patients were selected (0.898 in SELFI, 0.854 in ELISA, and 0.762 in LFIA, **** P < 0.0001). This indicates that SELFI is a more effective diagnostic system for the early diagnosis of PDAC than conventional ELISA. Our SELFI demonstrated exceptional convenience, sensitivity, and reliability, prompting us to propose a new approach for the early diagnosis of PDAC and its application to the preventive healthcare of other diseases. METHODS Materials Tetraethylorthosilicate (TEOS), (3-aminopropyl)triethoxysilane (APTES), gold (III) chloride trihydrate (HAuCl 4 ·3H₂O, 99.9%), sodium borohydride (NaBH 4 ), ascorbic acid, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC·HCl), N-hydroxysulfosuccinimide (sulfo-NHS), bovine serum albumin (BSA), 11-mercaptoundecanoic acid (11-MUA), 4-morpholineethanesulfonic acid (MES), AuNPs (20 nm diameter, 1 OD, stabilized suspension in citrate buffer), 96-well plate, and polyvinylpyrrolidone (PVP; Mw ~ 40,000) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Hydrochloric acid (HCl; 35–37%), ethanol (99%), potassium hydroxide (KOH), and buffer solution (pH 9.0 ± 0.02) were purchased from Samchun Chemical (Seoul, Republic of Korea). Ammonium hydroxide (NH 4 OH, 27%), sodium hydroxide (NaOH), and isopropyl alcohol (IPA) were purchased from Daejung (Busan, Republic of Korea). Phosphate-buffered saline (PBS; pH 7.4) and 0.5% (v/v) Tween 20 in PBS (PBST; pH 7.4) were purchased from Dyne Bio (Seongnam, Republic of Korea). Goat anti-mouse IgG antibody (Ab), backing card, nitrocellulose (NC) membrane, and absorbent pads were purchased from Bore Da Biotech Co., Ltd. (Seongnam, Republic of Korea). An ELISA kit for CA19-9 (Cat# EHCA199) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). CA19-9 monoclonal antibodies (A46300 and A46400) and CA19-9 antigen (J66100) were purchased from BiosPacific Inc. (Emeryville, CA, USA). Deionized water (D.W.) was used for all experiments. Instruments and Analysis TEM images of the NPs were obtained using a JEM-F200 instrument (JEOL, Akishima, Tokyo, Japan) at a maximum accelerated voltage of 200 kV. Cs-STEM images of the NPs were obtained using a JEM-ARM200F(NEOARM) (JEOL, Akishima, Tokyo, Japan). The UV-Vis-NIR extinction spectra of the NPs were obtained using an Optizen UV-Vis spectrometer (Mecasys, Daejeon, Republic of Korea). The solutions for the test and control lines were dispensed using an Automated Lateral Flow Reagent Dispenser (Claremont Bio, Upland, CA, USA). The colorimetric signal intensities of test line on NC membrane were measured by using ImageJ program. The colorimetric signal intensities of 96-microplate well after ELISA were measured by using BioTek Epoch 2 spectrophotometer (Agilent, Santa Clara, CA, USA). Numerical simulations were performed using COMSOL Multiphysics 5.4. Calibration curves for CA19-9 were obtained using OriginPro 8.5. The ROC curve and AUC values of each analysis system were calculated using homemade R program code with the pROC library 39 . Numerical Simulation of Nanostructures A wave optics simulation was performed to evaluate the optical behavior of the assembled nanostructure as a function of G Au . TE-polarized electromagnetic waves in the wavelength range of 400–800 nm was applied. The wavelength dependence of the refractive index of Au was obtained from the database of Werner 40 . Periodic conditions in the x- and y-directions and ports in the z-direction were applied for the backfield modeling. A perfectly matched layer in the x-, y-, and z-directions was set for scattered-field modeling. To build an assembled nanostructure with evenly distributed AuNPs, we placed an AuNP in contact with the SiO 2 NP and rotated it to a specific angle to form similar gaps, as observed in the experiment. Fabrication of SiO 2 NPs SiO 2 NPs of approximately 145 nm in diameter were fabricated using a modified Stöber method 41 . TEOS (1.5 mL) and NH 4 OH (1.5 mL) were added to absolute ethanol (35 mL) while stirring at 50°C and 700 rpm with a magnetic bar. After 2 h, the fabricated SiO 2 NPs were washed five times with ethanol via centrifugation at 8,500 rpm for 10 min. The concentration of SiO 2 NPs was adjusted to 50 mg mL - 1 with absolute ethanol. Amination of SiO 2 NPs (SiO 2 -NH 2 NPs) For the amination of SiO 2 NPs, 12.5 mg of SiO 2 NPs (in 250 µL of ethanol) was mixed with APTES (15.5 µL) and NH 4 OH (10 µL). The mixture was shaken overnight, after which the SiO 2 -NH₂ NPs were washed three times with ethanol via centrifugation at 8,500 rpm for 10 min. The concentration of the washed SiO 2 -NH 2 NPs was adjusted to 10 mg mL - 1 with absolute ethanol. Fabrication of Au Nanoseeds Prior to introduction, Au nanoseeds with diameters of 3–5 nm were prepared using the Martin method 42 . Briefly, 5.67 mg of NaBH 4 in a 50 mM NaOH solution (3 mL) and 19.6 mg of HAuCl 4 ∙3H 2 O in a 50 mM HCl solution (1 mL) were sequentially added to 96 mL of D.W. while stirring at 1,000 rpm. The color of the mixture changed to ruby-red after a short time, and the reaction continued for an additional 1 h. The prepared Au nanoseed mixture was stored at 4°C for 2–3 days before use. Fabrication of Au Nanoseed-Introduced SiO 2 NPs (SiO 2 @Au NPs) To fabricate the Au nanoseed-introduced SiO 2 NPs (SiO 2 @Au NPs), Au nanoseeds (10 mL) and PVP (20 mg) were mixed. To this mixture, 2 mg of SiO 2 -NH 2 NPs (in 200 µL of ethanol) was added, and the mixture was incubated overnight at room temperature. The fabricated SiO 2 @Au NPs were washed three times with ethanol via centrifugation at 8,500 rpm for 10 min. The washed SiO 2 @Au NPs were redispersed in 1 mL of 2% (w/v) PVP solution to adjust the concentration to 1 mg mL - 1 . Fabrication of AuNP-Assembled SiO 2 NPs (SiO 2 @Au@Au NPs) To fabricate the SiO 2 @Au@Au NPs, 0.2 mg of SiO 2 @Au was added to 9.8 mL of 2% (w/v) PVP solution, and the mixture was stirred at 500 rpm. While stirring, 20 µL of a 10 mM HAuCl 4 solution and 40 µL of a 10 mM ascorbic acid solution were added to the mixture every 5 min, with the number of injections cycles varying as 1, 3, 5, 7, 9, and 11 times. The mixture was stirred for an additional 10 min after the final addition. The fabricated SiO 2 @Au@Au NPs were washed five times with ethanol via centrifugation at 8,500 rpm for 10 min. Removal of SiO 2 NPs from SiO 2 @Au@Au 11 NPs First, 50 µg of SiO 2 @Au@Au 11 NPs was dispersed in 750 µL of IPA. Then, 250 µL of 4 M KOH was added, and the reaction was conducted for 24 h using an incubator mixer at 60°C and 300 rpm. After the reaction was complete, the supernatant was removed by centrifugation at 17,000 rpm for 15 min. The remaining AuNPs were sequentially washed twice with D.W. and 0.1% PBST (pH 7.4) via centrifugation at 17,000 rpm for 15 min. The washed AuNPs were redispersed in 1 mL of 0.1% PBST (pH 7.4). Conjugation of Antibodies onto SiO 2 @Au@Au 11 NPs Prior to the conjugation of antibodies, a carboxylic acid group was introduced by attaching 11-MUA onto SiO 2 @Au@Au 11 NPs. To achieve this, 0.1 mg of SiO 2 @Au@Au 11 NPs was redispersed in 90 µL of ethanol and mixed with 10 µL of a 2 mM 11-MUA solution. The mixture was shaken at room temperature for 1 h. After the reaction, the NPs were washed three times with ethanol via centrifugation at 8,500 rpm for 10 min. The surface-modified SiO 2 @Au@Au NPs were then redispersed in 100 µL of ethanol. Following the attachment of 11-MUA, the carboxylic acid group was activated. To activate the carboxylic acid group on the nanoprobes, a SiO 2 @Au@Au 11 NPs mixture (0.1 mg in 700 µL of D.W.) was sequentially mixed with 100 µL of a 2 mM EDC∙HCl solution, a 2 mM sulfo-NHS solution, and a 500 mM MES solution. The reaction mixture was shaken for 30 min, after which the supernatant was removed by centrifugation at 10,000 rpm for 10 min. The remaining nanoprobes were redispersed in 1 mL of 50 mM MES individually, and then 10 µL of a CA19-9 monoclonal Ab (A46400) solution (1 mg mL - 1 in pH 7.4 PBS) was added to each mixture. The mixture was shaken at room temperature for 2 h. The nanoprobes were washed twice with 50 mM MES via centrifugation at 10,000 rpm for 10 min and then redispersed in 1 mL of 50 mM MES after washing. After the reaction was complete, both nanoprobes were washed twice with 0.5% (v/v) PBST and 0.5% (w/v) BSA/PBS via centrifugation at 10,000 rpm for 10 min. The washed nanoprobes were stored at 4°C prior to use. Preparation of Ab-Attached Single AuNPs Prior to the attachment of antibodies onto the surface of single AuNPs, 1 mL of AuNP solution (6.54 × 10 11 particles mL - 1 , in citrate buffer) was removed via centrifugation, and the AuNPs were redispersed in 1 mL of buffer solution (pH 9.0). Then, 10 µL of CA19-9 monoclonal Ab (A46400) solution (1 mg mL - 1 , pH 7.4 PBS) was added. The mixture was shaken at room temperature for 30 min and stored in a refrigerator at 4°C for 24 h. Next, 100 µL of 10% (w/v) BSA/PBS was added to the mixture and shaken for 1 h. After the reaction was complete, the mixture was washed twice with PBS (pH 7.4) via centrifugation at 8,500 rpm for 10 min. The washed mixture was stored at 4°C before use. Preparation of Test Strips for CA19-9 Detection Test strips for SELFI and LFIA were prepared according to a previously published method with minor modifications 43 . After assembling the NC membrane on the backing card, 1 mg mL - 1 of goat anti-mouse IgG Ab solution was dispensed onto the control line. Subsequently, an anti-CA19-9 capture Ab solution (1 mg mL - 1 in pH 7.4 PBS) was dispensed onto the test line. The distance between the control and test lines was set at 4.0 mm. All solutions were dispensed onto the NC membrane at a flow rate of 1.9 µL cm - 1 . After dispensing, the NC membranes were dried in a desiccator for at least 2 h. After drying, the absorbent pad was attached to the backing card. The assembled strips were cut into 4 mm widths to fabricate individual test strips. Optimization of Analysis Time for SELFI Prior to conducting SELFI analysis, the solvent of the SiO 2 @Au@Au 11 NPs mixture was changed to 0.5% (v/v) PBST. A CA19-9 solution with a concentration of 370 U mL - 1 in PBS (pH 7.4) was prepared. As a developing mixture, 33 µL of a mixture consisting of 3 µL of CA19-9 solution, 5 µL of SiO 2 @Au@Au 11 NPs mixture, and 25 µL of 0.5% (v/v) PBST were mixed. The mixture was developed along each test strip for 1, 2, 5, 10, 15, 30, and 60 min. After development, the intensity of the colorimetric signal from the test strips was measured using the ImageJ software. Model Test for Detection of CA19-9 with Known Concentrations For SELFI and LFIA, the solvent of the SiO 2 @Au@Au 11 NPs mixture and single AuNPs was changed to 0.5% (v/v) PBST. Subsequently, CA19-9 solutions of known concentrations (0–200 U mL - 1 ) in PBS (pH 7.4) were prepared. Developing mixtures were prepared by mixing 3 µL of CA19-9 solution, 5 µL of SiO 2 @Au@Au 11 NPs or single AuNPs mixture, and 25 µL of 0.5% (v/v) PBST. The developed mixture was allowed to flow along the test strips for 15 min. The intensities of the colorimetric signals from the test lines were measured using the ImageJ software after capturing photographs of the NC membranes. For ELISA, the signal intensities of the prepared CA19-9 sample solutions were measured using a BioTek Epoch 2 spectrophotometer. Calibration Curves for Quantification of Biomarkers Using SELFI and LFIA Prior to the quantitative analysis of CA19-9, the solvent of the SiO 2 @Au@Au 11 NPs mixture and single AuNPs was changed to 0.5% (v/v) PBST. CA19-9 solutions with concentrations of 0, 0.0037, 0.037, 0.37, 3.7, 37, 370, 3700, and 37000 U mL - 1 in pH 7.4 PBS were prepared. After preparing developing mixtures by mixing 3 µL of CA19-9 solution, 5 µL of SiO 2 @Au@Au 11 NPs or single AuNPs mixture, and 25 µL of 0.5% (v/v) PBST, the developing mixtures were allowed to flow along the test strips for 15 min. The intensities of the colorimetric signals from the test lines were measured using the ImageJ software, and the calibration curve was calculated by sigmoidal fitting. After calculating the calibration curve, the LOD of each nanoprobe was determined based on the calibration curve 44 . For PSA and ICAM-1 as target biomarkers, the entire process remained the same, except that the sample solution was substituted with PSA solutions (at concentrations of 0, 0.004, 0.04, 0.4, 4, 40, 400, 4,000, and 40,000 ng mL - 1 in pH 7.4 PBS) or ICAM-1 solutions (at concentrations of 0, 0.002, 0.02, 0.2, 2, 20, 200, 2000, and 20000 ng mL - 1 in pH 7.4 PBS). Quantification of CA19-9 in Serum Samples Using ELISA The concentration of CA19-9 in serum samples was measured using ELISA kits for CA19-9 according to the manufacturer's instructions. Serum samples were measured in duplicate, starting with a four-fold dilution and continuing up to a 200-fold dilution. CA19-9 concentrations were determined from standard curves of control proteins from the kits using a four-parameter logistic nonlinear regression model with a BioTek Epoch 2 spectrophotometer. Quantification of CA19-9 in Serum Samples Using SELFI and LFIA Prior to the quantification of CA19-9 in the serum samples, the solvent of the nanoprobe mixture was changed to 0.5% (v/v) PBST. After that, 3 µL of serum samples, diluted 10-fold with 0.5% (v/v) PBST, were placed individually into a 96-well plate. The subsequent procedures were the same as those in the model test. The concentration of CA19-9 in each serum sample was calculated based on the signal intensity of the test lines and the calibration curves obtained in the model test. Statistical Analysis of Quantified CA19-9 Concentration Using ELISA, LFIA, and SELFI For the statistical analysis of quantified CA19-9 in the diagnosis of PDAC and comparison of each analytical system, an R program code incorporating the pROC library was designed and used. The total sample set consisted of three groups: 50 healthy controls, 40 patients with early-stage PDAC, and 60 patients with late-stage PDAC. To ensure robustness, 90% of the original samples (135 samples in this case) were selected by randomly excluding 10% of the samples from each group. The ROC curves and AUC values were generated from these 135 samples, and this process was repeated 100 times to obtain 100 ROC curves and corresponding AUC values. The average ROC curves for ELISA and SELFI were determined using the linear interpolation method, while the ROC curve for LFIA was obtained based on extreme value calculations. The average AUC values and 95% CI ranges were calculated from the 100 AUC values obtained. This process remained consistent across all other cases, except for variations in sample group composition. Declarations Code availability . Homemade R code for calculation of average ROC curves and AUC values are available on GitHub at https://github.com/JKim-45/SELFI. AUTHOR INFORMATION Corresponding Authors Jaehi Kim – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea; Email: [email protected] Jihwan Song – Department of Mechanical Engineering, Hanbat National University, Daejeon 34158, Republic of Korea; Email: [email protected] Jong-chan Lee – Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam 13620, Republic of Korea; College of Medicine, Seoul National University, Seoul 08826, Republic of Korea; Email: [email protected] Luke P. Lee – Renal Division and Division of Engineering in Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA 02115, USA.; Department of Bioengineering, Department of Electrical Engineering and Computer Science, University of California at Berkeley, Berkeley, CA 94720, USA.; Institute of Quantum Biophysics, Department of Biophysics, Sungkyunkwan University, Suwon 16419, Republic of Korea; Email: [email protected] Bong-Hyun Jun – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea; Email: [email protected] Authors ‡Sohyeon Jang – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea ‡Minsup Shin – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea ‡Jiseok Han – Department of Mechanical Engineering, Hanbat National University, Daejeon 34158, Republic of Korea Han-Joo Bae – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea Yuna Youn – Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam 13620, Republic of Korea Hye-Seong Cho – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea Kwanghee Yoo – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea Jun-Sik Chu – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea Jaehyun An – Company of BioSquare, Hwaseong 18449, Republic of Korea Hyejin Chang – Division of Science Education, Kangwon National University, Chuncheon 24341, Republic of Korea Jin-Hyeok Hwang – Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam 13620, Republic of Korea; College of Medicine, Seoul National University, Seoul 08826, Republic of Korea Author Contributions S.J., M.S., J.K., and B.-H.J. conceived and designed the experiments. S.J., M.S., J.H., H.-J.B., Y.Y, H.-S.C., K.Y., J.-S.C, J.A., and J.K. conducted experiments and analyses. S.J., M.S., J.H., Y.Y, J.K., J.-C.L., and B.-H.J. wrote the manuscript. H.C., J.-H.H., J.K., J.S., J.-C.L., L.P.L., and B.-H.J. revised the manuscript. L.P.L. and B.-H.J. supervised the overall process. All authors approved the final version of the manuscript. ‡These authors contributed equally. Funding Sources This study was funded by the Ministry of Science and ICT (grant number NRF-2022R1A2C2012883), Bio&Medical Technology Development Program, and a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (grant numbers 2021M3C1C3097211 and RS-2023-00222910).s Notes The authors declare no competing financial interests. References Siegel RL, Miller KD, Wagle NS, Jemal A (2023) Cancer statistics, 2023. CA Cancer J Clin 73:17–48 Mizrahi JD, Surana R, Valle JW, Shroff RT (2020) Pancreatic cancer. Lancet 395:2008–2020 Rawla P, Sunkara T, Gaduputi V (2019) Epidemiology of pancreatic cancer: Global trends, etiology and risk factors. World J Oncol 10:10–27 Rahib L et al (2014) Projecting cancer incidence and deaths to 2030: The unexpected burden of thyroid, liver, and pancreas cancers in the United States. Cancer Res 74:2913–2921 Kleeff J et al (2016) Pancreatic cancer. Nat Rev Dis Primers 2:16022 Neoptolemos JP et al (2018) Therapeutic developments in pancreatic cancer: current and future perspectives. Nat Rev Gastroenterol Hepatol 15:333–348 Yachida S, Iacobuzio-Donahue CA (2013) Evolution and dynamics of pancreatic cancer progression. Oncogene 32:5253–5260 Klos D et al (2010) Minimal residual disease in pancreatic cancer–Our first experiences. Rozhl Chir 89:135–139 Montagut C, Vidal J (2019) ctDNA to detect minimal residual disease in pancreatic cancer: Moving into clinical trials. Ann Oncol 30:1410–1413 Eissa MAL et al (2019) Promoter methylation of ADAMTS1 and BNC1 as potential biomarkers for early detection of pancreatic cancer in blood. Clin Epigenetics 11:59 Kobi M et al (2020) Imaging and management of pancreatic cancer. Semin Ultrasound CT MR 41:139–151 Kurihara K, Hanada K, Shimizu A (2020) Endoscopic ultrasonography diagnosis of early pancreatic cancer. Diagnostics 10:1086 Locker GY et al (2006) ASCO 2006 update of recommendations for the use of tumor markers in gastrointestinal cancer. J Clin Oncol 24:5313–5327 Ballehaninna UK, Chamberlain RS (2012) The clinical utility of serum CA 19 – 9 in the diagnosis, prognosis and management of pancreatic adenocarcinoma: An evidence based appraisal. J Gastrointest Oncol 3:105–119 McGuigan A et al (2018) Pancreatic cancer: A review of clinical diagnosis, epidemiology, treatment and outcomes. World J Gastroenterol 24:4846–4861 Blackford AL et al (2024) Pancreatic cancer surveillance and survival of high-risk individuals. JAMA Oncol 10:1087–1096 Cohen JD et al (2018) Detection and localization of surgically resectable cancers with a multi-analyte blood test. Science 359:926–930 Pereira SP et al (2020) Early detection of pancreatic cancer. Lancet Gastroenterol Hepatol 5:698–710 Di Nardo F, Chiarello M, Cavalera S, Baggiani C, Anfossi L (2021) Ten years of lateral flow immunoassay technique applications: Trends, challenges and future perspectives. Sensors 21:5185 Wang Z et al (2022) An overview for the nanoparticles-based quantitative lateral flow assay. Small Methods 6:2101143 Gao Z et al (2017) Platinum-decorated gold nanoparticles with dual functionalities for ultrasensitive colorimetric in vitro diagnostics. Nano Lett 17:5572–5579 Wei Z, Luciano K, Xia X (2022) Catalytic gold-iridium nanoparticles as labels for sensitive colorimetric lateral flow assay. ACS Nano 16:21609–21617 Kim H-M et al (2021) Multi-quantum dots-embedded silica-encapsulated nanoparticle-based lateral flow assay for highly sensitive exosome detection. Nanomaterials 11:768 Kim H-M et al (2021) Au–Ag assembled on silica nanoprobes for visual semiquantitative detection of prostate-specific antigen. J Nanobiotechnol 19:73 Bock S et al (2022) Lateral flow immunoassay with quantum-dot-embedded silica nanoparticles for prostate-specific antigen detection. Nanomaterials 12:33 Zhu J et al (2024) One-pot synthesized Au@Pt nanostars-based lateral flow immunoassay for colorimetric and photothermal dual-mode detection of SARS-CoV-2 nucleocapsid antibody. Anal Chim Acta 1292:342241 Wang Z et al (2024) Four-in-one multifunctional dandelion-like gold@platinum nanoparticles-driven multimodal lateral flow immunoassay. Small 20:2310869 Khelifa L, Hu Y, Jiang N, Yetisen AK (2022) Lateral flow assays for hormone detection. Lab Chip 22:2451–2475 Mahmoudi T, Shirdel B, Mansoori B, Baradaran B (2020) Dual sensitivity enhancement in gold nanoparticle-based lateral flow immunoassay for visual detection of carcinoembryonic antigen. Anal Sci Adv 1:161–172 Huang Y et al (2017) Lateral flow assay for carbohydrate antigen 19–9 in whole blood by using magnetized carbon nanotubes. Microchim Acta 184:4287–4294 Jiao X et al (2022) Lateral flow immunoassay based on time-resolved fluorescence microspheres for rapid and quantitative screening CA199 in human serum. Int J Mol Sci 23:9991 Shin M et al (2024) Highly sensitive multiplexed colorimetric lateral flow immunoassay by plasmon-controlled metal–silica isoform nanocomposites: PINs. Nano Converg 11:42 Kang H et al (2013) Near-infrared SERS nanoprobes with plasmonic Au/Ag hollow-shell assemblies for in vivo multiplex detection. Adv Funct Mater 23:3719–3727 Wang Y et al (2018) Establishment of time-resolved fluorescence immunochromatographic assay for detection of carbohydrate antigen 19 – 9. Sheng Wu Gong Cheng Xue Bao 34:1012–1018 Su C-W, Tian J-H, Ye J-J, Chang H-W, Tsai Y-C (2021) Construction of a label-free electrochemical immunosensor based on Zn-Co-S/graphene nanocomposites for carbohydrate antigen 19 – 9 detection. Nanomaterials 11:1475 Wu S et al (2024) Collard-like Bi2S3@ Au nanocomposites-based label free electrochemical immunosensor for quantitative detection of CA19-9. Talanta 127299 Zhao T, Jin B (2024) A label-free electrochemical biosensor based on a bimetallic organic framework for the detection of carbohydrate antigen 19 – 9. Anal Methods 16:6173–6182 Han X, Lin S, Li Y, Cheng C, Han X (2020) Near-infrared photothermal immunoassay for pancreatic cancer biomarker CA 19 – 9 on a digital thermometer. Anal Chim Acta 1098:117–124 Robin X et al (2011) pROC: an open-source package for R and S + to analyze and compare ROC curves. BMC Bioinform 12:77 Werner WSM, Glantschnig K, Ambrosch-Draxl C (2009) Optical constants and inelastic electronsScattering data for 17 elemental metals. J Phys Chem Ref Data 38:1013–1092 Stöber W, Fink A, Bohn E (1968) Controlled growth of monodisperse silica spheres in the micron size range. J Colloid Interface Sci 26:62–69 Martin MN, Basham JI, Chando P, Eah S-K (2010) Charged gold nanoparticles in non-polar solvents: 10-Min synthesis and 2D self-assembly. Langmuir 26:7410–7417 Bock S et al (2021) Lateral flow immunoassay with quantum-dot-embedded silica nanoparticles for prostate-specific antigen detection. Nanomaterials 12:33 Parolo C et al (2020) Tutorial: Design and fabrication of nanoparticle-based lateral-flow immunoassays. Nat Protoc 15:3788–3816 Additional Declarations There is NO Competing Interest. Supplementary Files Supporting.docx Supporting information Cite Share Download PDF Status: Published Journal Publication published 05 Feb, 2026 Read the published version in Nature Communications → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6695327","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":464055100,"identity":"f712d43d-5e4c-45b9-9813-eac99cb39f56","order_by":0,"name":"Bong-Hyun Jun","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAxklEQVRIiWNgGAWjYBACA4YDDAwfgFiCgQ0qQIwWxhkkamFgYOYhSYs54xmzxzZnbPIkZ6QlMPyoYTA2byCgxbLhjLlxzo20YmmJtAOMPccYzGQOEHLYgTNm0jkfDifOk0hvYOBtYLCRIOgXkBYLqBbGv0RrYbhxOHE20GHMQFvMCGqxbDhWJtlzJi1xZs+zhMMyxySMCWoxlzi8TeLHMZvEGcfTDB++qbExnEFIC4PEAQT7ACh6CAP+BiIUjYJRMApGwcgGAOGsQDJhtskWAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0001-6556-7946","institution":"Konkuk University","correspondingAuthor":true,"prefix":"","firstName":"Bong-Hyun","middleName":"","lastName":"Jun","suffix":""},{"id":464055101,"identity":"c8a23b45-626d-45bc-a016-67b5540b345b","order_by":1,"name":"Sohyeon Jang","email":"","orcid":"","institution":"Konkuk University","correspondingAuthor":false,"prefix":"","firstName":"Sohyeon","middleName":"","lastName":"Jang","suffix":""},{"id":464055102,"identity":"5a92cfa4-6e85-45c8-b0da-63091843fb6b","order_by":2,"name":"Minsup Shin","email":"","orcid":"https://orcid.org/0009-0001-6518-9401","institution":"Konkuk University","correspondingAuthor":false,"prefix":"","firstName":"Minsup","middleName":"","lastName":"Shin","suffix":""},{"id":464055103,"identity":"3c701968-242b-4aa2-b43a-b4187d4da61d","order_by":3,"name":"Jiseok Han","email":"","orcid":"","institution":"Hanbat National University","correspondingAuthor":false,"prefix":"","firstName":"Jiseok","middleName":"","lastName":"Han","suffix":""},{"id":464055104,"identity":"f6531cce-c765-415a-8f5b-8f2c1a8dcf3b","order_by":4,"name":"Han-Joo Bae","email":"","orcid":"","institution":"Konkuk University","correspondingAuthor":false,"prefix":"","firstName":"Han-Joo","middleName":"","lastName":"Bae","suffix":""},{"id":464055105,"identity":"0e1f14c6-0256-4b3a-9ce8-789994b631f1","order_by":5,"name":"Yuna Youn","email":"","orcid":"","institution":"Seoul National University Bundang Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yuna","middleName":"","lastName":"Youn","suffix":""},{"id":464055106,"identity":"a51f58cf-aeb8-4749-a3be-4226f9f64051","order_by":6,"name":"Hye-Seong Cho","email":"","orcid":"","institution":"Konkuk University","correspondingAuthor":false,"prefix":"","firstName":"Hye-Seong","middleName":"","lastName":"Cho","suffix":""},{"id":464055107,"identity":"1bd53509-2cda-4152-acc8-35adb5796f3d","order_by":7,"name":"Kwanghee Yoo","email":"","orcid":"https://orcid.org/0009-0004-8582-2165","institution":"Konkuk University","correspondingAuthor":false,"prefix":"","firstName":"Kwanghee","middleName":"","lastName":"Yoo","suffix":""},{"id":464055108,"identity":"db8d2e48-bbe8-47e1-8e61-a2dc05f3dd3f","order_by":8,"name":"Jun-Sik Chu","email":"","orcid":"","institution":"Konkuk University","correspondingAuthor":false,"prefix":"","firstName":"Jun-Sik","middleName":"","lastName":"Chu","suffix":""},{"id":464055109,"identity":"f6bb61df-86b8-4f1f-844a-e338107c0684","order_by":9,"name":"Jaehyun An","email":"","orcid":"","institution":"Company of BioSquare","correspondingAuthor":false,"prefix":"","firstName":"Jaehyun","middleName":"","lastName":"An","suffix":""},{"id":464055110,"identity":"585e88ed-efd9-44bc-b02f-a5a5f681e9bb","order_by":10,"name":"Hyejin Chang","email":"","orcid":"","institution":"Kangwon National University","correspondingAuthor":false,"prefix":"","firstName":"Hyejin","middleName":"","lastName":"Chang","suffix":""},{"id":464055111,"identity":"5538c460-a25d-4c81-9675-4dde774484c9","order_by":11,"name":"Jin-Hyeok Hwang","email":"","orcid":"","institution":"Seoul National University Bundang Hospital, Seoul National University College of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Jin-Hyeok","middleName":"","lastName":"Hwang","suffix":""},{"id":464055112,"identity":"f3c9b110-8636-4c22-9a0a-b2d6d08ea490","order_by":12,"name":"Jaehi Kim","email":"","orcid":"https://orcid.org/0000-0002-5150-4810","institution":"Konkuk University","correspondingAuthor":false,"prefix":"","firstName":"Jaehi","middleName":"","lastName":"Kim","suffix":""},{"id":464055113,"identity":"73470a01-eaa1-4efb-91a6-36f7d25369b4","order_by":13,"name":"Jihwan Song","email":"","orcid":"https://orcid.org/0000-0003-1711-3418","institution":"Hanbat National University","correspondingAuthor":false,"prefix":"","firstName":"Jihwan","middleName":"","lastName":"Song","suffix":""},{"id":464055114,"identity":"eb8166eb-3320-469e-aacf-5c23fc10fafc","order_by":14,"name":"Jong-chan Lee","email":"","orcid":"","institution":"Seoul National University Bundang Hospital","correspondingAuthor":false,"prefix":"","firstName":"Jong-chan","middleName":"","lastName":"Lee","suffix":""},{"id":464055115,"identity":"c8bcf27f-f57d-4fde-8c42-39ace74d359e","order_by":15,"name":"Luke Lee","email":"","orcid":"","institution":"Harvard Medical School","correspondingAuthor":false,"prefix":"","firstName":"Luke","middleName":"","lastName":"Lee","suffix":""}],"badges":[],"createdAt":"2025-05-19 05:45:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6695327/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6695327/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41467-026-69204-7","type":"published","date":"2026-02-05T05:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":83750851,"identity":"5c1ff45c-e1fe-4abf-9f45-cce9f160cc49","added_by":"auto","created_at":"2025-06-02 06:46:44","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":808589,"visible":true,"origin":"","legend":"\u003cp\u003eQuantification analysis of CA19-9 for early diagnosis of PDAC using the SELFI system. \u003cstrong\u003ea\u003c/strong\u003e, \u0026nbsp;Illustration for pancreas and PDAC cells. PDAC cells are located in pancreas which located at deep anatomical location. Overexpressed CA19-9 by PDAC cells is released into the bloodstream and elevate CA19-9 level in serum sample. \u003cstrong\u003eb\u003c/strong\u003e, Illustration for the assembled structure of nanoprobe. Fabricated nanoprobes had number of hotspots with nano-gap which amplifying the colorimetric signal. \u003cstrong\u003ec\u003c/strong\u003e, Schematic illustration of SELFI system for PDAC diagnosis. Colorimetric signal at test line was exhibited by binding of nanoprobes at test line of strip via interaction between CA19-9 and CA19-9 antibody. \u003cstrong\u003ed\u003c/strong\u003e, Comparison of the SELFI system with ELISA, one of the conventional methods for biomarker quantification analysis. The SELFI system required only 15 min for analysis, whereas ELISA required 285 min. \u003cstrong\u003ee\u003c/strong\u003e, Changes of E-field norm and colorimetric signal intensity of nanoprobes according to the gap distance between AuNPs. As the gap distance between AuNPs was being closer, E-field norm increased via generation of strong hotspots. \u003cstrong\u003ef\u003c/strong\u003e, Model test for CA19-9 detection using LFIA, in which single 20 nm AuNPs were used as colorimetric nanoprobes, and SELFI. Compared with single AuNP-based LFIA, SELFI, which usescolorimetric nanoprobes with an assembled nanostructure, exhibitedincreased signal intensity and a wider dynamic range. \u003cstrong\u003eg\u003c/strong\u003e, Schematic illustration of the ROC curve for PDAC diagnosis using ELISA (orange line), LFIA (red line), and SELFI (blue line). The gray dotted line represents a random classifier.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6695327/v1/2d921ea4a5e71f489c0b15cd.png"},{"id":83751285,"identity":"d4bcb6b9-4f82-45bf-80b6-5bc6e9e11a52","added_by":"auto","created_at":"2025-06-02 06:54:44","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1227767,"visible":true,"origin":"","legend":"\u003cp\u003eDesign with numerical simulation, fabrication, and characterization of colorimetric nanoprobes for SELFI. \u003cstrong\u003ea\u003c/strong\u003e, Maximum electric field amplification of the assembled nanostructure induced by hotspots between AuNPs on SiO\u003csub\u003e2\u003c/sub\u003e NP. \u003cstrong\u003eb\u003c/strong\u003e, Electric field distribution at peak wavelength depending on \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e. \u003cstrong\u003ec\u003c/strong\u003e, Extinction cross-section spectra depending on \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e. \u003cstrong\u003ed\u003c/strong\u003e, Comparison of electric field norm distribution (at each peak wavelength) between a single AuNP and an AuNP on the assembled nanostructure depending on AuNP diameter. \u003cstrong\u003ee\u003c/strong\u003e, Extinction cross-section intensity of a single AuNP (red line) and an AuNP on the assembled nanostructure (blue line) at peak wavelength depending on AuNP diameter. \u003cstrong\u003ef\u003c/strong\u003e, Schematic illustration of fabricated nanoprobes with an assembled nanostructure. AuNPs were located onto the surface of SiO\u003csub\u003e2\u003c/sub\u003e nanotemplates with controlled nanogaps. \u003cstrong\u003eg\u003c/strong\u003e, Cs-STEM images of fabricated SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e1\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e3\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e5\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e7\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e9\u003c/sub\u003e, and SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e. The average gap distances between AuNPs were 9.55 nm, 7.63 nm, 5.89 nm, 3.53 nm, 1.33 nm, and 1.20 nm, respectively (n = 20). Scale bars, 50 nm (upper images) and 20 nm (lower images). \u003cstrong\u003eh\u003c/strong\u003e, UV-Vis extinction spectra of fabricated NPs. \u003cstrong\u003ei\u003c/strong\u003e, Colorimetric signal intensities of fabricated NPs on NC membrane (n = 3). The colorimetric signal intensity was measured using the ImageJ software. \u003cstrong\u003ej\u003c/strong\u003e, UV-Vis extinction spectra of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e (blue solid line), AuNPs from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e (blue dotted line), and 3–5 nm AuNPs (red dotted line). \u003cstrong\u003ek\u003c/strong\u003e, Extinction at λ\u003csub\u003emax.ext\u003c/sub\u003e of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e (blue filled bar), AuNPs from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e (blue hatched bar), and 3–5 nm AuNPs (red hatched bar) (n = 3).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6695327/v1/80ec49d81353f825e453ab4d.png"},{"id":83750853,"identity":"48c04c2d-496c-470a-a508-29a1d288e885","added_by":"auto","created_at":"2025-06-02 06:46:44","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":412855,"visible":true,"origin":"","legend":"\u003cp\u003eModel test for quantification of CA19-9 using SELFI. \u003cstrong\u003ea\u003c/strong\u003e, Schematic illustration of time-dependent signal intensity from ELISA, conventional LFIA, and SELFI during biomarker quantification analysis. \u003cstrong\u003eb\u003c/strong\u003e, (i) Photographs of the NC membrane after CA19-9 quantification analysis depending on the analysis time. (ii) Signal intensity graph as a function of analysis time (n = 3). The concentration of CA19-9 in the analyte sample was 3,700 U mL\u003csup\u003e-1\u003c/sup\u003e, and the calibration curve was fitted using the Boltzmann model. \u003cstrong\u003ec\u003c/strong\u003e, (i) Photographs of the NC membrane after developing CA19-9 solution with SELFI and LFIA, depending on the concentration of CA19-9. (ii) Signal intensity graph as a function of CA19-9 concentration (n = 3). The blue line, red line, and orange dotted line represent signals from SELFI, LFIA, and ELISA, respectively. The gray dotted line represents the boundary for naked-eye detection in SELFI and LFIA. \u003cstrong\u003ed\u003c/strong\u003e, Calibration curve for CA19-9 quantification (n = 3). \u003cstrong\u003ee\u003c/strong\u003e, Calibration curve for PSA quantification (n = 3). \u003cstrong\u003ef\u003c/strong\u003e, Calibration curve for ICAM1 quantification (n = 3). \u003cstrong\u003eg\u003c/strong\u003e, Comparison of SELFI for PDAC diagnosis with other LFIA-based diagnostic systems (1–2) and other diagnostic systems including ELISA (3–6). For \u003cstrong\u003eb–f\u003c/strong\u003e, the colorimetric signal intensity was measured using the ImageJ software. For \u003cstrong\u003ed–f\u003c/strong\u003e, the blue line and circles represent the results using SELFI, while the red line and circles represent the results using LFIA. For \u003cstrong\u003ed–f\u003c/strong\u003e, the calibration curves were fitted using sigmoidal curves.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6695327/v1/c2c9c8d532d1d2d7f4d234cb.png"},{"id":83750849,"identity":"a6e9e31e-659b-4c52-b5a4-1985b8e21d67","added_by":"auto","created_at":"2025-06-02 06:46:44","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":343825,"visible":true,"origin":"","legend":"\u003cp\u003eDiagnosis of PDAC with clinical serum samples using ELISA, LFIA, and SELFI. \u003cstrong\u003ea\u003c/strong\u003e, Schematic illustration of the statistical data analysis process. \u003cstrong\u003eb\u003c/strong\u003e, Average ROC curve for ELISA, LFIA, and SELFI using samples from early-stage PDAC patients and healthy controls (AUC\u003csub\u003eELISA\u003c/sub\u003e = 0.854, AUC\u003csub\u003eLFIA\u003c/sub\u003e = 0.762, AUC\u003csub\u003eSELFI\u003c/sub\u003e = 0.898). \u003cstrong\u003ec,\u003c/strong\u003e Average ROC curve for ELISA, LFIA, and SELFI using samples from late-stage PDAC patients and healthy controls (AUC\u003csub\u003eELISA\u003c/sub\u003e = 0.926, AUC\u003csub\u003eLFIA\u003c/sub\u003e = 0.834, AUC\u003csub\u003eSELFI\u003c/sub\u003e = 0.925). \u003cstrong\u003ed\u003c/strong\u003e, Average ROC curve for ELISA, LFIA, and SELFI using total samples (AUC\u003csub\u003eELISA\u003c/sub\u003e = 0.896, AUC\u003csub\u003eLFIA\u003c/sub\u003e = 0.805, AUC\u003csub\u003eSELFI\u003c/sub\u003e = 0.913). \u003cstrong\u003ee\u003c/strong\u003e, Violin plot of AUC values for ELISA, LFIA, and SELFI using samples from early-stage PDAC patients and healthy controls (****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001). \u003cstrong\u003ef\u003c/strong\u003e, Violin plot of AUC values for ELISA, LFIA, and SELFI using samples from late-stage PDAC patients and healthy controls (****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001; ns \u003cem\u003eP\u003c/em\u003e = 0.3535). \u003cstrong\u003eg\u003c/strong\u003e, Violin plot of AUC values for ELISA, LFIA, and SELFI using total samples (****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001). For \u003cstrong\u003eb–g\u003c/strong\u003e, data from ELISA, LFIA, and SELFI are depicted in orange, red, and blue, respectively. The ROC curves and AUC values were obtained using a homemade R program code with the pROC library. For \u003cstrong\u003eb–d,\u003c/strong\u003e the average ROC curves were obtained by averaging 100 ROC curves using linear interpolation, with the black dotted line representing a random classifier. For \u003cstrong\u003ee–g\u003c/strong\u003e, the median value is depicted as a solid line, and the interquartile range boundaries are depicted as dotted lines in each violin plot.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6695327/v1/0582224d7f8919c606c49b3d.png"},{"id":105445996,"identity":"6adfabdd-603f-4db1-bb29-8033eb421359","added_by":"auto","created_at":"2026-03-26 07:12:59","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3906473,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6695327/v1/820f4978-c6e7-4f5f-bd49-535103e0b4f1.pdf"},{"id":83750855,"identity":"aa011b05-9898-432d-ae8d-12b54546929e","added_by":"auto","created_at":"2025-06-02 06:46:44","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":6359104,"visible":true,"origin":"","legend":"Supporting information","description":"","filename":"Supporting.docx","url":"https://assets-eu.researchsquare.com/files/rs-6695327/v1/ca4c90f5c95354872cfaeb9d.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Early Diagnosis of Pancreatic Ductal Adenocarcinoma by Signal-Enhanced Lateral Flow Immunoassay: SELFI","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003ePancreatic ductal adenocarcinoma (PDAC) is one of the deadliest solid tumors, characterized by a persistently poor prognosis\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e. Over the past few decades, the incidence of PDAC has steadily increased; however, the 5-year survival rate has remained below 10%\u003csup\u003e1, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. This grim prognosis is primarily due to three factors: difficulties in early detection, the aggressive biology of the tumor, and resistance to current therapies\u003csup\u003e\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u003c/sup\u003e. Among various approaches to overcoming these obstacles, early detection and curative resection are the most critical and effective strategies\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u003c/sup\u003e. Furthermore, even in postoperative patients, early detection of minimal residual disease or small recurrent tumors is essential for timely intervention\u003csup\u003e\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eDespite its importance, early diagnosis of PDAC is challenging because significant symptoms are not observed in the early stages, and the pancreas is difficult to examine using standard diagnostic methods due to its deep anatomical location\u003csup\u003e\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. PDAC is currently diagnosed using pancreatic imaging techniques such as computed tomography, magnetic resonance imaging, and endoscopic ultrasound (EUS) examinations\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e\u003c/sup\u003e. If a pancreatic tumor is not identified using these methods, invasive diagnostic approaches, such as EUS-guided fine needle aspiration (EUS-FNA), may be required to confirm the presence of a tumor\u003csup\u003e\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. However, these complex and invasive procedures can be burdensome for patients, leading to reluctance and delays in diagnosis, resulting in missed optimal diagnostic opportunities. Therefore, alternative diagnostic methods that are more convenient and accessible compared to existing techniques are essential for the early diagnosis of PDAC.\u003c/p\u003e \u003cp\u003eAs a noninvasive approach, diagnosing pancreatic ductal adenocarcinoma (PDAC) using biomarkers in biological samples such as serum has been proposed. Numerous biomarkers for PDAC detection have been investigated; however, cancer antigen 19\u0026thinsp;\u0026minus;\u0026thinsp;9 (CA19-9) remains the only FDA-approved biomarker\u003csup\u003e\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). CA19-9 has demonstrated reliability in evaluating treatment responses in advanced PDAC; however, it has limitations as a screening marker for early-stage resectable PDAC when used with existing biomarker screening systems, such as enzyme-linked immunosorbent assay (ELISA)\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u003c/sup\u003e. To overcome the limitations of CA19-9, research efforts like CancerSEEK have aimed to discover new blood biomarkers beyond CA19-9\u003csup\u003e17, 18\u003c/sup\u003e. Despite these initiatives, the discovery of novel PDAC biomarkers presents challenges related to reproducibility and commercial viability.\u003c/p\u003e \u003cp\u003eAs an alternative, the development of a novel quantification system for CA19-9 to replace ELISA has been explored to enhance the sensitivity and efficacy of early PDAC diagnosis. One candidate method for CA19-9 quantification in PDAC diagnosis is the lateral flow immunoassay (LFIA), which is widely used for point-of-care testing due to its convenience. Generally, the LFIA system produces red-colored signals from gold nanoparticles (AuNPs) used as probes. The LFIA system is more user-friendly than conventional liquid biopsy methods, and the time necessary to obtain analysis results is very short\u003csup\u003e\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u003c/sup\u003e. However, the colorimetric signal intensity of the AuNP probe is insufficient, necessitating the development of a new probe\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u003c/sup\u003e. Consequently, recent reports have shown that not only silver nanoparticles (NPs) but also alloy-type metal NPs\u0026mdash;such as gold/platinum and gold/iridium\u0026mdash;and even NPs with assembled nanostructures, in which quantum dots or metal NPs are assembled onto silica NPs, have been utilized as probes for LFIA\u003csup\u003e\u003cspan additionalcitationids=\"CR22 CR23 CR24 CR25 CR26\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. Due to its great convenience, LFIA has been successfully applied to analyze various targets, such as exosomes\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e, prostate cancer antigen\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e, hormones\u003csup\u003e\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e, carcinoembryonic antigen\u003csup\u003e\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e, and the COVID-19 virus\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. However, the quantitative analysis of CA19-9 using LFIA for early PDAC diagnosis has rarely been reported, and in the few existing studies, the limit of detection (LOD), which reflects the sensitivity of the analytical system, was not adequate for highly sensitive analysis\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn this study, we designed a Signal-Enhanced Lateral Flow Immunoassay (SELFI) system for the early diagnosis of PDAC through the quantitative analysis of CA19-9 in serum samples. Silica nanoparticles (SiO\u003csub\u003e2\u003c/sub\u003e NPs) with assembled gold nanoparticles (AuNPs) (SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs), which feature numerous hotspots due to nanogaps between the assembled AuNPs, were fabricated as the colorimetric nanoprobes in the SELFI system (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e1\u003c/span\u003eb and c). Our SELFI system required only 15 minutes after sample loading for the quantitative analysis of CA19-9 in serum samples, while the commonly used ELISA for biomarker quantification requires over 4 hours and involves a complex experimental process (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e1\u003c/span\u003ed). The colorimetric signal intensity of SELFI was significantly higher than that of conventional lateral flow immunoassays (LFIA) at the same CA19-9 concentration, thanks to the hotspots in SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs, which markedly enhanced colorimetric signal intensity (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e1\u003c/span\u003ee). The limit of detection (LOD) of CA19-9 in SELFI was 27.6-fold lower than that of conventional LFIA (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e1\u003c/span\u003ef). As a diagnostic system for PDAC based on the quantitative analysis of CA19-9, SELFI exhibited superior diagnostic performance compared to both ELISA and conventional LFIA, as indicated by their receiver operating characteristic (ROC) curve and area under the curve (AUC) values. Notably, SELFI demonstrated significant improvement in the early diagnosis of PDAC (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e1\u003c/span\u003eg).\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eDesign of Colorimetric Nanoprobe for SELFI Using Numerical Simulation\u003c/h2\u003e \u003cp\u003eBased on previous studies, hotspots form between metal NPs, such as AuNPs or Ag/Au NPs, when they assemble on the surface of SiO\u003csub\u003e2\u003c/sub\u003e NPs (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ea)\u003csup\u003e\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u003c/sup\u003e. However, the relationship between the intensity of the electric field\u0026mdash;enhanced by the hotspot effect\u0026mdash;and parameters like the gap distance between the AuNPs (\u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e) and the diameter of the AuNPs (\u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e) has yet to be explored. Numerical simulations were performed to investigate the relationships between the optical properties of AuNPs-assembled silica NPs and these parameters, particularly \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e and \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e, to aid in the \u003cem\u003ein silico\u003c/em\u003e design of innovative colorimetric nanoprobes for SELFI.\u003c/p\u003e \u003cp\u003eSupplementary Fig.\u0026nbsp;1 illustrates a schematic of the 3D model geometry and the various cases considered in the simulation. The assembled nanostructure comprised SiO\u003csub\u003e2\u003c/sub\u003e NPs with a diameter of 168 nm, and 270 AuNPs were evenly distributed on its surface. The electric field was oriented in the y-direction. Six simulation cases were analyzed, with \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e values of 10, 8, 6, 4, 2, and 1 nm. Since the diameter of the SiO\u003csub\u003e2\u003c/sub\u003e NP and the count of AuNPs were fixed, \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e increased as \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e decreased. Using these designed assembled nanostructures, a wave-optics simulation was performed in the visible wavelength range, assuming that the assembled nanostructures were immersed in water.\u003c/p\u003e \u003cp\u003eBased on the simulation results, the electric field norm spectra of each model in the visible wavelength range were analyzed (Supplementary Fig.\u0026nbsp;2a). The electric field on the surface of the assembled nanostructure became stronger as \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e decreased, and hotspots emerged between the AuNPs when \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e was less than 6 nm (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003eb). The results indicated that the overall intensity of the electric field increased with decreasing \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e, with a particularly notable increase at the peak wavelength. The extinction cross-sectional spectra, which represent the sum of the absorption and scattering cross-sections, also increased as \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e decreased and exhibited a peak at a wavelength similar to that of the electric field norm spectra (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ec, Supplementary Fig.\u0026nbsp;2b, c).\u003c/p\u003e \u003cp\u003eAlthough the extinction cross-section increased as \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e decreased, it remained unclear whether this increase was attributed to the reduction in \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e or the rise in \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e, as both parameters changed simultaneously. To determine which parameter had a more significant impact on increasing the extinction cross-section of the nanostructure, the change in the extinction cross-section of a single AuNP on the assembled nanostructure was compared with that of a single AuNP on a SiO\u003csub\u003e2\u003c/sub\u003e substrate (Supplementary Figs.\u0026nbsp;3 and 4, Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ed, e). As shown in Supplementary Fig.\u0026nbsp;4, the electric field of a single AuNP on the SiO\u003csub\u003e2\u003c/sub\u003e substrate increased as \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e increased. In contrast, the electric field from the AuNPs on the assembled nanostructure increased dramatically with the creation of hotspots between AuNPs when \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e was reduced to a certain distance (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ed). Due to this phenomenon, the extinction cross-section of AuNPs in the assembled nanostructure was nearly seven times higher than that of a single AuNP on the SiO\u003csub\u003e2\u003c/sub\u003e substrate at the largest \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e, despite being nearly identical at the smallest \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ee). This comparison suggests that \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e, which is closely linked to the generation of hotspots between AuNPs, plays a more significant role in enhancing the electric field than \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e.\u003c/p\u003e \u003cp\u003eWe confirmed that the assembled nanostructure exhibited a significantly higher extinction intensity than a single 20 nm AuNP, the gold standard nanoprobe for conventional LFIA (Supplementary Fig.\u0026nbsp;5). These simulation results suggest that the amplification of the electric field around the assembled nanostructure, resulting in a strong extinction intensity, is not solely due to the increase in \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e but is mainly driven by the formation of hotspots caused by the decrease in \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e as \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e increases.\u003c/p\u003e \u003cp\u003eBased on these simulation results, we could expect that our designed SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs should exhibit a more intense colorimetric signal due to the electric field enhancement effect caused by hotspots formed between AuNPs with under the specific distance. This enhancement renders them superior to single AuNPs, which are commonly used as nanoprobes in conventional LFIA system.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eFabrication and Characterization of Nanoprobes\u003c/h3\u003e\n\u003cp\u003eBased on the simulation results, we fabricated SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs via a seed-mediated growth method, with controlling the nanogaps as colorimetric nanoprobes for SELFI (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ef and Supplementary Fig.\u0026nbsp;6). AuNPs were grown after introducing 3\u0026ndash;5 nm AuNPs onto the surface of aminated SiO\u003csub\u003e2\u003c/sub\u003e NPs, with control over the amount of Au precursor and ascorbic acid. Each particle was named as SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e1\u003c/sub\u003e to SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e according to the reagents used.\u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003eg shows the Cs-corrected scanning transmission electron microscopy (TEM) (Cs-STEM) images of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e1\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e3\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e5\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e7\u003c/sub\u003e, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e9\u003c/sub\u003e, and SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e. The average \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e for each nanostructure was 9.55, 7.63, 5.89, 3.53, 1.33, and 1.20 nm, respectively. This decrease in \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e resulted from the increase in \u003cem\u003eD\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e, which was based on the amount of Au\u003csup\u003e3+\u003c/sup\u003e precursor used during the fabrication process (Supplementary Fig.\u0026nbsp;8).\u003c/p\u003e \u003cp\u003eAfter fabricating SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs, the optical properties of each nanoprobe were characterized to select the optimal colorimetric nanoprobes for SELFI. First, the degree of visible light extinction, strongly related to the colorimetric signal intensity of the SELFI system, was measured. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003eh, the extinction of visible light by the SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs increased as \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs was decreased, following the same trend observed in the numerical simulation. Based on these extinction spectra, the color of the SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NP mixture deepened as \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs was decreased (Supplementary Fig.\u0026nbsp;9). Similarly, the intensity of the colorimetric signals from the SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs on the NC membrane exhibited the same trend, with the colorimetric signal intensity from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs nearly reaching the maximum measurable value (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ei).\u003c/p\u003e \u003cp\u003eTo identify the experimental primary factor behind this increase, the extinction spectra of Au nanoseeds (3\u0026ndash;5 nm), SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs, and AuNPs obtained from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs after removing the SiO\u003csub\u003e2\u003c/sub\u003e nanotemplate were measured (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ej). The effect of AuNP diameter was evaluated by comparing the extinction of Au nanoseeds with that of single AuNPs from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs, which could be regarded as individual AuNPs with large interparticle distances. The effect of gap distance was assessed by comparing SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs with AuNPs from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs, as the AuNP diameter remained unchanged after SiO\u003csub\u003e2\u003c/sub\u003e nanotemplate removal.\u003c/p\u003e \u003cp\u003eA comparison of the maximum extinction of each nanostructure at their λ\u003csub\u003emax.ext\u003c/sub\u003e showed that the degree of extinction increased slightly from 0.031 (Au nanoseeds) to 0.120 (single AuNPs from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs) due to the increase in AuNP diameter. However, the extinction increased significantly to 0.426 for SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs, attributed to the localized surface plasmon resonance effect from the nanogaps between the AuNPs (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e2\u003c/span\u003ek). These results experimentally confirm that the extinction of visible light by the fabricated nanostructure is more influenced by the gap distance between AuNPs than by AuNP diameter.\u003c/p\u003e \u003cp\u003eTo verify the superior optical properties of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs as colorimetric nanoprobes, they were compared with 20 nm AuNPs for conventional LFIA, and 200 nm AuNPs, which are similar in diameter to SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs. In terms of visible light extinction, the maximum extinction value of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs was 58-fold higher than that of 200 nm AuNPs and 1,439-fold higher than that of 20 nm AuNPs (Supplementary Fig.\u0026nbsp;10). As shown in Supplementary Fig.\u0026nbsp;11, when each type of nanoprobe was observed using dark-field microscopy, only SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs appeared brightly due to the surface plasmon resonance effect originating from the hotspots of the assembled nanostructure, whereas the others, which lacked hotspots, appeared faintly.\u003c/p\u003e \u003cp\u003eFinally, the colorimetric signal intensity of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs on the NC membrane was significantly higher than that of 20 nm AuNPs on the NC membrane when the particle concentration exceeded a certain level (Supplementary Fig.\u0026nbsp;12). In summary, our fabricated SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs exhibit superior optical properties as colorimetric nanoprobes compared to conventional AuNPs.\u003c/p\u003e\n\u003ch3\u003eQuantification of CA19-9 Using SELFI\u003c/h3\u003e\n\u003cp\u003eAfter fabricating and characterizing SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs, a model test was conducted for the quantitative analysis of CA19-9 using the SELFI system, which has the potential to be faster than ELISA and more sensitive than LFIA (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e3\u003c/span\u003ea). First, the change in colorimetric signal over time was examined to determine the optimal analysis time for SELFI. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e3\u003c/span\u003eb, the intensity of the test line's colorimetric signal increased with time, with no further significant changes observable by the naked eye after 15 minutes. This trend was further confirmed by quantifying the colorimetric signal intensity using ImageJ, which demonstrated that the intensity saturated after 15 minutes. These results indicate that quantification of CA19-9 in the sample using SELFI can be completed in just 15 min.\u0026mdash;significantly shorter than the 285 minutes required for ELISA.\u003c/p\u003e \u003cp\u003eNext, the colorimetric signals obtained from the SELFI, LFIA, and ELISA systems at various CA19-9 concentrations (0\u0026ndash;200 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) were compared (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e3\u003c/span\u003ec). In the SELFI system, represented by a navy-colored line as a signal, the colorimetric signal of test line signal was detectable by the naked eye when the CA19-9 concentration exceeded 5 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. However, no visible signals were detected in the LFIA system until the CA19-9 concentration reached 100 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, and even at this concentration, only a faint red line was observed. These tendencies were further confirmed by quantifying the colorimetric signal intensity using the ImageJ software. For the SELFI system, the signal intensity was 6.34 at a CA19-9 concentration of 1 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, increasing 14-fold to 90.2 at 200 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e. This increase was significantly greater than that observed in the LFIA system, where the signal intensity increased only two-fold, from 8.81 to 19.15, over the same CA19-9 concentration range. Considering that ELISA, a well-established high-sensitivity detection method for target biomarkers, exhibited a similar signal increase pattern, these results suggest that the SELFI system offers a sensitivity comparable to that of ELISA for the detection of target biomarkers, including CA19-9.\u003c/p\u003e \u003cp\u003eAfter verifying the fundamental characteristics of the SELFI system, calibration curves were obtained for the quantification of CA19-9 using each nanoprobe and corresponding standard solutions. The colorimetric signals from the test lines of both nanoprobes were measured using the ImageJ software, and the calibration curve of each analysis method was determined through sigmoidal fitting (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e3\u003c/span\u003ed and Supplementary Fig.\u0026nbsp;14). The equation for the calibration curve for SELFI was as follows:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:Signal\\:intensity=83.84+\\frac{(14.75-83.84)}{1+{\\left(\\frac{Concentration}{20.62}\\right)}^{0.85}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eFor single AuNP-based LFIA, the calibration curve was:\u003cdiv id=\"Equb\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equb\" name=\"EquationSource\"\u003e\n$$\\:Signal\\:intensity=31.13+\\frac{(6.45-31.13)}{1+{\\left(\\frac{Concentration}{689.82}\\right)}^{1.26}}$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eThe LOD of each system was calculated based on these calibration curves. The LOD for CA19-9 using SELFI was 0.15 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e, which was 27.6-fold lower than that of single AuNP-based LFIA (4.14 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) and two-fold lower than that of commercially available ELISA (0.3 U mL\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e). Furthermore, the LOD of SELFI was lower than that of previously reported LFIA systems for CA19-9 quantification\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, indicating that SELFI is among the most sensitive LFIA systems for CA19-9 detection. Similar trends were observed not only for CA19-9 but also for prostate-specific antigen (PSA) and intercellular adhesion molecule 1 (ICAM-1), which are frequently used as cancer biomarkers. Therefore, the SELFI system could serve as a substitute for conventional LFIA systems, regardless of the target biomarker type (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e3\u003c/span\u003ee, f).\u003c/p\u003e \u003cp\u003eTo further assess the performance of SELFI as a diagnostic system, CA19-9 quantification-based PDAC diagnostic methods were compared by plotting analysis time versus LOD (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e3\u003c/span\u003eg). The results demonstrated that SELFI quantification of CA19-9 required a short analysis time, comparable to conventional LFIA\u003csup\u003e\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e, LFIA with carbon nanotubes (1)\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e, and time-resolved fluorescence microsphere-based LFIA (2)\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e, all of which have relatively high LODs. Conversely, SELFI exhibited a lower LOD than or similar to that of commercially available ELISA kits, electrochemical immunosensors (3, 4, and 5)\u003csup\u003e\u003cspan additionalcitationids=\"CR36\" citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e\u003c/sup\u003e, and near-infrared photothermal immunoassays (6)\u003csup\u003e\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e\u003c/sup\u003e, which require longer analysis time. These results highlight that SELFI effectively combines the advantages of different diagnostic systems, providing both rapid analysis and a low LOD.\u003c/p\u003e\n\u003ch3\u003eDiagnosis of PDAC via CA19-9 Quantification in Serum Samples Using SELFI\u003c/h3\u003e\n\u003cp\u003eUsing LFIA and the designed SELFI system, quantification of CA19-9 in serum samples from healthy controls (50 ea.), patients in early-stage PDAC (40 ea.), and patients in late-stage PDAC (60 ea.) was conducted to determine the possibility of early diagnosis of PDAC using the SELFI system. After taking photographs of the test strips, the concentration of CA19-9 in each serum sample was successfully calculated by measuring the colorimetric signal intensity of the test lines using the ImageJ software (Supplementary Fig.\u0026nbsp;15, 16, and Supplementary Table S2). To compare the diagnostic performance of SELFI with that of ELISA, which is commonly used for the quantification of CA19-9, and LFIA, the ROC curve and AUC values of each system were compared. The ROC curve and AUC value were calculated after 90% of the samples were extracted from the mother populations, and this process was repeated 100 times to enhance statistical reliability (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003ea).\u003c/p\u003e \u003cp\u003eFigure\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003eb\u0026ndash;d shows the average ROC curves of each system, in which the mother population was set as samples from healthy controls and early-stage patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003eb), healthy controls and late-stage patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003ec), and total samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003ed). The average AUC value of SELFI was significantly higher than that of ELISA for samples from healthy controls and early-stage patients (0.898 and 0.854, respectively), indicating that our SELFI system performed notably better for the early diagnosis of PDAC than ELISA. The AUC values of both diagnostic systems were almost the same for samples from healthy controls and late-stage patients (0.925 for SELFI and 0.926 for ELISA, respectively). For the total serum samples, the AUC value of SELFI was slightly higher than that of ELISA (0.913 for SELFI and 0.896 for ELISA, respectively). Finally, LFIA, which used 20 nm AuNPs as colorimetric nanoprobes, showed poor AUC values in all cases compared to the other two systems (0.762 for healthy controls and early-stage patients, 0.834 for healthy controls and late-stage patients, and 0.805 for total samples). The calculated AUC values for all repetition steps were plotted as violin plots to confirm the statistical evaluation of the differences in AUC values among ELISA, LFIA, and SELFI (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e4\u003c/span\u003ee\u0026ndash;g). \u003cem\u003eP\u003c/em\u003e values were less than 0.0001, except in the case of SELFI and ELISA for healthy controls and late-stage patients (0.3535), and the 95% confidence intervals (CIs) of the AUC values from SELFI were narrow (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Based on these results, the superior performance of the SELFI system for PDAC diagnosis, especially for early diagnosis, was statistically proven.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAUC calculations for ELISA, LFIA, and SELFI results at 95% CIs (100 repetitions)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eSamples\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"6\" nameend=\"c7\" namest=\"c2\"\u003e \u003cp\u003eDiagnosis system\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eELISA\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e \u003cp\u003eLFIA\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c7\" namest=\"c6\"\u003e \u003cp\u003eSELFI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eAUC\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003e95% CI\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEarly-stage (n\u0026thinsp;=\u0026thinsp;40)\u003c/p\u003e \u003cp\u003e+ Healthy control (n\u0026thinsp;=\u0026thinsp;50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.854\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.834\u0026ndash;0.881\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.762\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.736\u0026ndash;0.792\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.898\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.878\u0026ndash;0.924\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLate-stage (n\u0026thinsp;=\u0026thinsp;60)\u003c/p\u003e \u003cp\u003e+ Healthy control (n\u0026thinsp;=\u0026thinsp;50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.917\u0026ndash;0.940\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.834\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.815\u0026ndash;0.861\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.925\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.914\u0026ndash;0.949\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEarly-stage (n\u0026thinsp;=\u0026thinsp;40)\u003c/p\u003e \u003cp\u003e\u0026thinsp;+\u0026thinsp;Late-stage (n\u0026thinsp;=\u0026thinsp;60)\u003c/p\u003e \u003cp\u003e+ Healthy control (n\u0026thinsp;=\u0026thinsp;50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.896\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.885\u0026ndash;0.913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.805\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.789\u0026ndash;0.822\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e0.900\u0026ndash;0.931\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e"},{"header":"CONCLUSION","content":"\u003cp\u003eWe designed and fabricated SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs, where Au NPs were embedded onto the surface of SiO\u003csub\u003e2\u003c/sub\u003e NPs, as colorimetric nanoprobes for the LFIA system for the early diagnosis of PDAC. The SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs demonstrated more intense colorimetric signals than the commercially used single Au NPs, due to the plasmonic enhancement effect from the nanogaps between the Au NPs. Based on the most intense colorimetric signal observed at the closest gap distance, SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs were selected as the optimal nanoprobes among the fabricated NPs. Similar to other LFIA systems, the quantification of CA19-9 was performed within 15 minutes, and LOD of SELFI was 0.15 U mL\u003csup\u003e− 1\u003c/sup\u003e, which was 27.6-fold lower than that of single AuNPs. The concentration of CA19-9 in serum samples was determined after SELFI, and the AUC value of the ROC curve from SELFI was 0.913, which was higher than the values of 0.896 from ELISA and 0.805 from LFIA (****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001). Notably, the differences in AUC values between SELFI-ELISA and SELFI-LFIA became more pronounced when samples from healthy controls and early-stage patients were selected (0.898 in SELFI, 0.854 in ELISA, and 0.762 in LFIA, ****\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001). This indicates that SELFI is a more effective diagnostic system for the early diagnosis of PDAC than conventional ELISA. Our SELFI demonstrated exceptional convenience, sensitivity, and reliability, prompting us to propose a new approach for the early diagnosis of PDAC and its application to the preventive healthcare of other diseases.\u003c/p\u003e "},{"header":"METHODS","content":"\u003ch2\u003eMaterials\u003c/h2\u003e\u003cp\u003eTetraethylorthosilicate (TEOS), (3-aminopropyl)triethoxysilane (APTES), gold (III) chloride trihydrate (HAuCl\u003csub\u003e4\u003c/sub\u003e·3H₂O, 99.9%), sodium borohydride (NaBH\u003csub\u003e4\u003c/sub\u003e), ascorbic acid, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (EDC·HCl), N-hydroxysulfosuccinimide (sulfo-NHS), bovine serum albumin (BSA), 11-mercaptoundecanoic acid (11-MUA), 4-morpholineethanesulfonic acid (MES), AuNPs (20 nm diameter, 1 OD, stabilized suspension in citrate buffer), 96-well plate, and polyvinylpyrrolidone (PVP; Mw ~ 40,000) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Hydrochloric acid (HCl; 35–37%), ethanol (99%), potassium hydroxide (KOH), and buffer solution (pH 9.0 ± 0.02) were purchased from Samchun Chemical (Seoul, Republic of Korea). Ammonium hydroxide (NH\u003csub\u003e4\u003c/sub\u003eOH, 27%), sodium hydroxide (NaOH), and isopropyl alcohol (IPA) were purchased from Daejung (Busan, Republic of Korea). Phosphate-buffered saline (PBS; pH 7.4) and 0.5% (v/v) Tween 20 in PBS (PBST; pH 7.4) were purchased from Dyne Bio (Seongnam, Republic of Korea).\u003c/p\u003e\u003cp\u003eGoat anti-mouse IgG antibody (Ab), backing card, nitrocellulose (NC) membrane, and absorbent pads were purchased from Bore Da Biotech Co., Ltd. (Seongnam, Republic of Korea). An ELISA kit for CA19-9 (Cat# EHCA199) was purchased from Thermo Fisher Scientific (Waltham, MA, USA). CA19-9 monoclonal antibodies (A46300 and A46400) and CA19-9 antigen (J66100) were purchased from BiosPacific Inc. (Emeryville, CA, USA). Deionized water (D.W.) was used for all experiments.\u003c/p\u003e\u003ch3\u003eInstruments and Analysis\u003c/h3\u003e\u003cp\u003eTEM images of the NPs were obtained using a JEM-F200 instrument (JEOL, Akishima, Tokyo, Japan) at a maximum accelerated voltage of 200 kV. Cs-STEM images of the NPs were obtained using a JEM-ARM200F(NEOARM) (JEOL, Akishima, Tokyo, Japan). The UV-Vis-NIR extinction spectra of the NPs were obtained using an Optizen UV-Vis spectrometer (Mecasys, Daejeon, Republic of Korea).\u003c/p\u003e\u003cp\u003eThe solutions for the test and control lines were dispensed using an Automated Lateral Flow Reagent Dispenser (Claremont Bio, Upland, CA, USA). The colorimetric signal intensities of test line on NC membrane were measured by using ImageJ program. The colorimetric signal intensities of 96-microplate well after ELISA were measured by using BioTek Epoch 2 spectrophotometer (Agilent, Santa Clara, CA, USA).\u003c/p\u003e\u003cp\u003eNumerical simulations were performed using COMSOL Multiphysics 5.4. Calibration curves for CA19-9 were obtained using OriginPro 8.5. The ROC curve and AUC values of each analysis system were calculated using homemade R program code with the pROC library\u003csup\u003e\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003ch2\u003eNumerical Simulation of Nanostructures\u003c/h2\u003e\u003cp\u003eA wave optics simulation was performed to evaluate the optical behavior of the assembled nanostructure as a function of \u003cem\u003eG\u003c/em\u003e\u003csub\u003eAu\u003c/sub\u003e. TE-polarized electromagnetic waves in the wavelength range of 400–800 nm was applied. The wavelength dependence of the refractive index of Au was obtained from the database of Werner\u003csup\u003e\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e\u003c/sup\u003e. Periodic conditions in the x- and y-directions and ports in the z-direction were applied for the backfield modeling. A perfectly matched layer in the x-, y-, and z-directions was set for scattered-field modeling.\u003c/p\u003e\u003cp\u003eTo build an assembled nanostructure with evenly distributed AuNPs, we placed an AuNP in contact with the SiO\u003csub\u003e2\u003c/sub\u003e NP and rotated it to a specific angle to form similar gaps, as observed in the experiment.\u003c/p\u003e\u003ch2\u003eFabrication of SiO\u003csub\u003e2\u003c/sub\u003e NPs\u003c/h2\u003e\u003cp\u003eSiO\u003csub\u003e2\u003c/sub\u003e NPs of approximately 145 nm in diameter were fabricated using a modified Stöber method\u003csup\u003e\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e\u003c/sup\u003e. TEOS (1.5 mL) and NH\u003csub\u003e4\u003c/sub\u003eOH (1.5 mL) were added to absolute ethanol (35 mL) while stirring at 50°C and 700 rpm with a magnetic bar. After 2 h, the fabricated SiO\u003csub\u003e2\u003c/sub\u003e NPs were washed five times with ethanol via centrifugation at 8,500 rpm for 10 min. The concentration of SiO\u003csub\u003e2\u003c/sub\u003e NPs was adjusted to 50 mg mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e with absolute ethanol.\u003c/p\u003e\u003ch2\u003eAmination of SiO\u003csub\u003e2\u003c/sub\u003e NPs (SiO\u003csub\u003e2\u003c/sub\u003e-NH\u003csub\u003e2\u003c/sub\u003e NPs)\u003c/h2\u003e\u003cp\u003eFor the amination of SiO\u003csub\u003e2\u003c/sub\u003e NPs, 12.5 mg of SiO\u003csub\u003e2\u003c/sub\u003e NPs (in 250 µL of ethanol) was mixed with APTES (15.5 µL) and NH\u003csub\u003e4\u003c/sub\u003eOH (10 µL). The mixture was shaken overnight, after which the SiO\u003csub\u003e2\u003c/sub\u003e-NH₂ NPs were washed three times with ethanol via centrifugation at 8,500 rpm for 10 min. The concentration of the washed SiO\u003csub\u003e2\u003c/sub\u003e-NH\u003csub\u003e2\u003c/sub\u003e NPs was adjusted to 10 mg mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e with absolute ethanol.\u003c/p\u003e\u003ch2\u003eFabrication of Au Nanoseeds\u003c/h2\u003e\u003cp\u003ePrior to introduction, Au nanoseeds with diameters of 3–5 nm were prepared using the Martin method\u003csup\u003e\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e\u003c/sup\u003e. Briefly, 5.67 mg of NaBH\u003csub\u003e4\u003c/sub\u003e in a 50 mM NaOH solution (3 mL) and 19.6 mg of HAuCl\u003csub\u003e4\u003c/sub\u003e∙3H\u003csub\u003e2\u003c/sub\u003eO in a 50 mM HCl solution (1 mL) were sequentially added to 96 mL of D.W. while stirring at 1,000 rpm. The color of the mixture changed to ruby-red after a short time, and the reaction continued for an additional 1 h. The prepared Au nanoseed mixture was stored at 4°C for 2–3 days before use.\u003c/p\u003e\u003ch2\u003eFabrication of Au Nanoseed-Introduced SiO\u003csub\u003e2\u003c/sub\u003e NPs (SiO\u003csub\u003e2\u003c/sub\u003e@Au NPs)\u003c/h2\u003e\u003cp\u003eTo fabricate the Au nanoseed-introduced SiO\u003csub\u003e2\u003c/sub\u003e NPs (SiO\u003csub\u003e2\u003c/sub\u003e@Au NPs), Au nanoseeds (10 mL) and PVP (20 mg) were mixed. To this mixture, 2 mg of SiO\u003csub\u003e2\u003c/sub\u003e-NH\u003csub\u003e2\u003c/sub\u003e NPs (in 200 µL of ethanol) was added, and the mixture was incubated overnight at room temperature. The fabricated SiO\u003csub\u003e2\u003c/sub\u003e@Au NPs were washed three times with ethanol via centrifugation at 8,500 rpm for 10 min. The washed SiO\u003csub\u003e2\u003c/sub\u003e@Au NPs were redispersed in 1 mL of 2% (w/v) PVP solution to adjust the concentration to 1 mg mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003ch2\u003eFabrication of AuNP-Assembled SiO\u003csub\u003e2\u003c/sub\u003e NPs (SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs)\u003c/h2\u003e\u003cp\u003eTo fabricate the SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs, 0.2 mg of SiO\u003csub\u003e2\u003c/sub\u003e@Au was added to 9.8 mL of 2% (w/v) PVP solution, and the mixture was stirred at 500 rpm. While stirring, 20 µL of a 10 mM HAuCl\u003csub\u003e4\u003c/sub\u003e solution and 40 µL of a 10 mM ascorbic acid solution were added to the mixture every 5 min, with the number of injections cycles varying as 1, 3, 5, 7, 9, and 11 times. The mixture was stirred for an additional 10 min after the final addition. The fabricated SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs were washed five times with ethanol via centrifugation at 8,500 rpm for 10 min.\u003c/p\u003e\u003ch2\u003eRemoval of SiO\u003csub\u003e2\u003c/sub\u003e NPs from SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs\u003c/h2\u003e\u003cp\u003eFirst, 50 µg of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs was dispersed in 750 µL of IPA. Then, 250 µL of 4 M KOH was added, and the reaction was conducted for 24 h using an incubator mixer at 60°C and 300 rpm. After the reaction was complete, the supernatant was removed by centrifugation at 17,000 rpm for 15 min. The remaining AuNPs were sequentially washed twice with D.W. and 0.1% PBST (pH 7.4) via centrifugation at 17,000 rpm for 15 min. The washed AuNPs were redispersed in 1 mL of 0.1% PBST (pH 7.4).\u003c/p\u003e\u003ch2\u003eConjugation of Antibodies onto SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs\u003c/h2\u003e\u003cp\u003ePrior to the conjugation of antibodies, a carboxylic acid group was introduced by attaching 11-MUA onto SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs. To achieve this, 0.1 mg of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs was redispersed in 90 µL of ethanol and mixed with 10 µL of a 2 mM 11-MUA solution. The mixture was shaken at room temperature for 1 h. After the reaction, the NPs were washed three times with ethanol via centrifugation at 8,500 rpm for 10 min. The surface-modified SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au NPs were then redispersed in 100 µL of ethanol.\u003c/p\u003e\u003cp\u003eFollowing the attachment of 11-MUA, the carboxylic acid group was activated. To activate the carboxylic acid group on the nanoprobes, a SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs mixture (0.1 mg in 700 µL of D.W.) was sequentially mixed with 100 µL of a 2 mM EDC∙HCl solution, a 2 mM sulfo-NHS solution, and a 500 mM MES solution. The reaction mixture was shaken for 30 min, after which the supernatant was removed by centrifugation at 10,000 rpm for 10 min. The remaining nanoprobes were redispersed in 1 mL of 50 mM MES individually, and then 10 µL of a CA19-9 monoclonal Ab (A46400) solution (1 mg mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e in pH 7.4 PBS) was added to each mixture. The mixture was shaken at room temperature for 2 h.\u003c/p\u003e\u003cp\u003eThe nanoprobes were washed twice with 50 mM MES via centrifugation at 10,000 rpm for 10 min and then redispersed in 1 mL of 50 mM MES after washing. After the reaction was complete, both nanoprobes were washed twice with 0.5% (v/v) PBST and 0.5% (w/v) BSA/PBS via centrifugation at 10,000 rpm for 10 min. The washed nanoprobes were stored at 4°C prior to use.\u003c/p\u003e\u003ch2\u003ePreparation of Ab-Attached Single AuNPs\u003c/h2\u003e\u003cp\u003ePrior to the attachment of antibodies onto the surface of single AuNPs, 1 mL of AuNP solution (6.54 × 10\u003csup\u003e11\u003c/sup\u003e particles mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e, in citrate buffer) was removed via centrifugation, and the AuNPs were redispersed in 1 mL of buffer solution (pH 9.0). Then, 10 µL of CA19-9 monoclonal Ab (A46400) solution (1 mg mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e, pH 7.4 PBS) was added. The mixture was shaken at room temperature for 30 min and stored in a refrigerator at 4°C for 24 h. Next, 100 µL of 10% (w/v) BSA/PBS was added to the mixture and shaken for 1 h. After the reaction was complete, the mixture was washed twice with PBS (pH 7.4) via centrifugation at 8,500 rpm for 10 min. The washed mixture was stored at 4°C before use.\u003c/p\u003e\u003ch2\u003ePreparation of Test Strips for CA19-9 Detection\u003c/h2\u003e\u003cp\u003eTest strips for SELFI and LFIA were prepared according to a previously published method with minor modifications\u003csup\u003e\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e\u003c/sup\u003e. After assembling the NC membrane on the backing card, 1 mg mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e of goat anti-mouse IgG Ab solution was dispensed onto the control line. Subsequently, an anti-CA19-9 capture Ab solution (1 mg mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e in pH 7.4 PBS) was dispensed onto the test line. The distance between the control and test lines was set at 4.0 mm. All solutions were dispensed onto the NC membrane at a flow rate of 1.9 µL cm\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e. After dispensing, the NC membranes were dried in a desiccator for at least 2 h. After drying, the absorbent pad was attached to the backing card. The assembled strips were cut into 4 mm widths to fabricate individual test strips.\u003c/p\u003e\u003ch2\u003eOptimization of Analysis Time for SELFI\u003c/h2\u003e\u003cp\u003ePrior to conducting SELFI analysis, the solvent of the SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs mixture was changed to 0.5% (v/v) PBST. A CA19-9 solution with a concentration of 370 U mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e in PBS (pH 7.4) was prepared. As a developing mixture, 33 µL of a mixture consisting of 3 µL of CA19-9 solution, 5 µL of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs mixture, and 25 µL of 0.5% (v/v) PBST were mixed. The mixture was developed along each test strip for 1, 2, 5, 10, 15, 30, and 60 min. After development, the intensity of the colorimetric signal from the test strips was measured using the ImageJ software.\u003c/p\u003e\u003ch2\u003eModel Test for Detection of CA19-9 with Known Concentrations\u003c/h2\u003e\u003cp\u003eFor SELFI and LFIA, the solvent of the SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs mixture and single AuNPs was changed to 0.5% (v/v) PBST. Subsequently, CA19-9 solutions of known concentrations (0–200 U mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e) in PBS (pH 7.4) were prepared. Developing mixtures were prepared by mixing 3 µL of CA19-9 solution, 5 µL of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs or single AuNPs mixture, and 25 µL of 0.5% (v/v) PBST. The developed mixture was allowed to flow along the test strips for 15 min. The intensities of the colorimetric signals from the test lines were measured using the ImageJ software after capturing photographs of the NC membranes. For ELISA, the signal intensities of the prepared CA19-9 sample solutions were measured using a BioTek Epoch 2 spectrophotometer.\u003c/p\u003e\u003ch2\u003eCalibration Curves for Quantification of Biomarkers Using SELFI and LFIA\u003c/h2\u003e\u003cp\u003ePrior to the quantitative analysis of CA19-9, the solvent of the SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs mixture and single AuNPs was changed to 0.5% (v/v) PBST. CA19-9 solutions with concentrations of 0, 0.0037, 0.037, 0.37, 3.7, 37, 370, 3700, and 37000 U mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e in pH 7.4 PBS were prepared. After preparing developing mixtures by mixing 3 µL of CA19-9 solution, 5 µL of SiO\u003csub\u003e2\u003c/sub\u003e@Au@Au\u003csub\u003e11\u003c/sub\u003e NPs or single AuNPs mixture, and 25 µL of 0.5% (v/v) PBST, the developing mixtures were allowed to flow along the test strips for 15 min. The intensities of the colorimetric signals from the test lines were measured using the ImageJ software, and the calibration curve was calculated by sigmoidal fitting. After calculating the calibration curve, the LOD of each nanoprobe was determined based on the calibration curve\u003csup\u003e\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e\u003cp\u003eFor PSA and ICAM-1 as target biomarkers, the entire process remained the same, except that the sample solution was substituted with PSA solutions (at concentrations of 0, 0.004, 0.04, 0.4, 4, 40, 400, 4,000, and 40,000 ng mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e in pH 7.4 PBS) or ICAM-1 solutions (at concentrations of 0, 0.002, 0.02, 0.2, 2, 20, 200, 2000, and 20000 ng mL\u003csup\u003e-\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u003c/sup\u003e in pH 7.4 PBS).\u003c/p\u003e\u003ch2\u003eQuantification of CA19-9 in Serum Samples Using ELISA\u003c/h2\u003e\u003cp\u003eThe concentration of CA19-9 in serum samples was measured using ELISA kits for CA19-9 according to the manufacturer's instructions. Serum samples were measured in duplicate, starting with a four-fold dilution and continuing up to a 200-fold dilution. CA19-9 concentrations were determined from standard curves of control proteins from the kits using a four-parameter logistic nonlinear regression model with a BioTek Epoch 2 spectrophotometer.\u003c/p\u003e\u003ch2\u003eQuantification of CA19-9 in Serum Samples Using SELFI and LFIA\u003c/h2\u003e\u003cp\u003ePrior to the quantification of CA19-9 in the serum samples, the solvent of the nanoprobe mixture was changed to 0.5% (v/v) PBST. After that, 3 µL of serum samples, diluted 10-fold with 0.5% (v/v) PBST, were placed individually into a 96-well plate. The subsequent procedures were the same as those in the model test. The concentration of CA19-9 in each serum sample was calculated based on the signal intensity of the test lines and the calibration curves obtained in the model test.\u003c/p\u003e\u003ch2\u003eStatistical Analysis of Quantified CA19-9 Concentration Using ELISA, LFIA, and SELFI\u003c/h2\u003e\u003cp\u003eFor the statistical analysis of quantified CA19-9 in the diagnosis of PDAC and comparison of each analytical system, an R program code incorporating the pROC library was designed and used. The total sample set consisted of three groups: 50 healthy controls, 40 patients with early-stage PDAC, and 60 patients with late-stage PDAC. To ensure robustness, 90% of the original samples (135 samples in this case) were selected by randomly excluding 10% of the samples from each group.\u003c/p\u003e\u003cp\u003eThe ROC curves and AUC values were generated from these 135 samples, and this process was repeated 100 times to obtain 100 ROC curves and corresponding AUC values. The average ROC curves for ELISA and SELFI were determined using the linear interpolation method, while the ROC curve for LFIA was obtained based on extreme value calculations. The average AUC values and 95% CI ranges were calculated from the 100 AUC values obtained. This process remained consistent across all other cases, except for variations in sample group composition.\u003c/p\u003e"},{"header":"Declarations","content":"\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e.\u003c/p\u003e\n\u003cp\u003eHomemade R code for calculation of average ROC curves and AUC values are available on GitHub at https://github.com/JKim-45/SELFI.\u003c/p\u003e\n\u003cp\u003eAUTHOR INFORMATION\u003c/p\u003e\n\u003cp\u003eCorresponding Authors\u003c/p\u003e\n\u003cp\u003eJaehi Kim – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea; Email: [email protected]\u003c/p\u003e\n\u003cp\u003eJihwan Song – Department of Mechanical Engineering, Hanbat National University, Daejeon 34158, Republic of Korea; Email: [email protected]\u003c/p\u003e\n\u003cp\u003eJong-chan Lee –\u0026nbsp;Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam\u0026nbsp;13620,\u0026nbsp;Republic of\u0026nbsp;Korea;\u0026nbsp;College of Medicine, Seoul National University, Seoul 08826, Republic of Korea; Email: [email protected]\u003c/p\u003e\n\u003cp\u003eLuke P. Lee – Renal Division and Division of Engineering in Medicine, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA 02115, USA.; Department of Bioengineering, Department of Electrical Engineering and Computer Science, University of California at Berkeley, Berkeley, CA 94720, USA.; Institute of Quantum Biophysics, Department of Biophysics, Sungkyunkwan University, Suwon 16419, Republic of Korea; Email:\u0026nbsp;[email protected]\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBong-Hyun Jun – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea; Email: [email protected]\u003c/p\u003e\n\u003cp\u003eAuthors\u003c/p\u003e\n\u003cp\u003e‡Sohyeon Jang – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea\u003c/p\u003e\n\u003cp\u003e‡Minsup Shin – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea\u003c/p\u003e\n\u003cp\u003e‡Jiseok Han – Department of Mechanical Engineering, Hanbat National University, Daejeon 34158, Republic of Korea\u003c/p\u003e\n\u003cp\u003eHan-Joo Bae – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea\u003c/p\u003e\n\u003cp\u003eYuna Youn –\u0026nbsp;Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam\u0026nbsp;13620,\u0026nbsp;Republic of\u0026nbsp;Korea\u003c/p\u003e\n\u003cp\u003eHye-Seong Cho\u0026nbsp;– Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea\u003c/p\u003e\n\u003cp\u003eKwanghee Yoo – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea\u003c/p\u003e\n\u003cp\u003eJun-Sik Chu – Department of Bioscience and Biotechnology, Konkuk University, Seoul 05029, Republic of Korea\u003c/p\u003e\n\u003cp\u003eJaehyun An – Company of BioSquare, Hwaseong 18449, Republic of Korea\u003c/p\u003e\n\u003cp\u003eHyejin Chang – Division of Science Education, Kangwon National University, Chuncheon 24341, Republic of Korea\u003c/p\u003e\n\u003cp\u003eJin-Hyeok Hwang –\u0026nbsp;Department of Internal Medicine, Seoul National University Bundang Hospital, Seongnam\u0026nbsp;13620,\u0026nbsp;Republic of\u0026nbsp;Korea;\u0026nbsp;College of Medicine, Seoul National University, Seoul 08826, Republic of Korea\u003c/p\u003e\n\u003cp\u003eAuthor Contributions\u003c/p\u003e\n\u003cp\u003eS.J., M.S., J.K., and B.-H.J. conceived and designed the experiments. S.J., M.S., J.H., H.-J.B., Y.Y, H.-S.C., K.Y., J.-S.C, J.A., and J.K. conducted experiments and analyses. S.J., M.S., J.H., Y.Y, J.K., J.-C.L., and B.-H.J. wrote the manuscript. H.C., J.-H.H., J.K., J.S., J.-C.L., L.P.L., and B.-H.J. revised the manuscript. L.P.L. and B.-H.J. supervised the overall process. All authors approved the final version of the manuscript.\u003cbr\u003e\u0026nbsp;‡These authors contributed equally.\u003c/p\u003e\n\u003cp\u003eFunding Sources\u003c/p\u003e\n\u003cp\u003eThis study was funded by the Ministry of Science and ICT (grant number NRF-2022R1A2C2012883), Bio\u0026amp;Medical Technology Development Program, and a National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (grant numbers 2021M3C1C3097211 and RS-2023-00222910).s\u003c/p\u003e\n\u003cp\u003eNotes\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing financial interests.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSiegel RL, Miller KD, Wagle NS, Jemal A (2023) Cancer statistics, 2023. CA Cancer J Clin 73:17\u0026ndash;48\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMizrahi JD, Surana R, Valle JW, Shroff RT (2020) Pancreatic cancer. Lancet 395:2008\u0026ndash;2020\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRawla P, Sunkara T, Gaduputi V (2019) Epidemiology of pancreatic cancer: Global trends, etiology and risk factors. World J Oncol 10:10\u0026ndash;27\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRahib L et al (2014) Projecting cancer incidence and deaths to 2030: The unexpected burden of thyroid, liver, and pancreas cancers in the United States. Cancer Res 74:2913\u0026ndash;2921\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKleeff J et al (2016) Pancreatic cancer. Nat Rev Dis Primers 2:16022\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNeoptolemos JP et al (2018) Therapeutic developments in pancreatic cancer: current and future perspectives. Nat Rev Gastroenterol Hepatol 15:333\u0026ndash;348\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYachida S, Iacobuzio-Donahue CA (2013) Evolution and dynamics of pancreatic cancer progression. Oncogene 32:5253\u0026ndash;5260\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlos D et al (2010) Minimal residual disease in pancreatic cancer\u0026ndash;Our first experiences. Rozhl Chir 89:135\u0026ndash;139\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMontagut C, Vidal J (2019) ctDNA to detect minimal residual disease in pancreatic cancer: Moving into clinical trials. Ann Oncol 30:1410\u0026ndash;1413\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEissa MAL et al (2019) Promoter methylation of ADAMTS1 and BNC1 as potential biomarkers for early detection of pancreatic cancer in blood. Clin Epigenetics 11:59\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKobi M et al (2020) Imaging and management of pancreatic cancer. Semin Ultrasound CT MR 41:139\u0026ndash;151\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKurihara K, Hanada K, Shimizu A (2020) Endoscopic ultrasonography diagnosis of early pancreatic cancer. Diagnostics 10:1086\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLocker GY et al (2006) ASCO 2006 update of recommendations for the use of tumor markers in gastrointestinal cancer. J Clin Oncol 24:5313\u0026ndash;5327\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBallehaninna UK, Chamberlain RS (2012) The clinical utility of serum CA 19\u0026thinsp;\u0026ndash;\u0026thinsp;9 in the diagnosis, prognosis and management of pancreatic adenocarcinoma: An evidence based appraisal. J Gastrointest Oncol 3:105\u0026ndash;119\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcGuigan A et al (2018) Pancreatic cancer: A review of clinical diagnosis, epidemiology, treatment and outcomes. World J Gastroenterol 24:4846\u0026ndash;4861\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBlackford AL et al (2024) Pancreatic cancer surveillance and survival of high-risk individuals. JAMA Oncol 10:1087\u0026ndash;1096\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCohen JD et al (2018) Detection and localization of surgically resectable cancers with a multi-analyte blood test. Science 359:926\u0026ndash;930\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePereira SP et al (2020) Early detection of pancreatic cancer. Lancet Gastroenterol Hepatol 5:698\u0026ndash;710\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDi Nardo F, Chiarello M, Cavalera S, Baggiani C, Anfossi L (2021) Ten years of lateral flow immunoassay technique applications: Trends, challenges and future perspectives. Sensors 21:5185\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z et al (2022) An overview for the nanoparticles-based quantitative lateral flow assay. Small Methods 6:2101143\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGao Z et al (2017) Platinum-decorated gold nanoparticles with dual functionalities for ultrasensitive colorimetric in vitro diagnostics. Nano Lett 17:5572\u0026ndash;5579\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWei Z, Luciano K, Xia X (2022) Catalytic gold-iridium nanoparticles as labels for sensitive colorimetric lateral flow assay. ACS Nano 16:21609\u0026ndash;21617\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim H-M et al (2021) Multi-quantum dots-embedded silica-encapsulated nanoparticle-based lateral flow assay for highly sensitive exosome detection. Nanomaterials 11:768\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim H-M et al (2021) Au\u0026ndash;Ag assembled on silica nanoprobes for visual semiquantitative detection of prostate-specific antigen. J Nanobiotechnol 19:73\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBock S et al (2022) Lateral flow immunoassay with quantum-dot-embedded silica nanoparticles for prostate-specific antigen detection. Nanomaterials 12:33\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu J et al (2024) One-pot synthesized Au@Pt nanostars-based lateral flow immunoassay for colorimetric and photothermal dual-mode detection of SARS-CoV-2 nucleocapsid antibody. Anal Chim Acta 1292:342241\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Z et al (2024) Four-in-one multifunctional dandelion-like gold@platinum nanoparticles-driven multimodal lateral flow immunoassay. Small 20:2310869\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhelifa L, Hu Y, Jiang N, Yetisen AK (2022) Lateral flow assays for hormone detection. Lab Chip 22:2451\u0026ndash;2475\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMahmoudi T, Shirdel B, Mansoori B, Baradaran B (2020) Dual sensitivity enhancement in gold nanoparticle-based lateral flow immunoassay for visual detection of carcinoembryonic antigen. Anal Sci Adv 1:161\u0026ndash;172\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang Y et al (2017) Lateral flow assay for carbohydrate antigen 19\u0026ndash;9 in whole blood by using magnetized carbon nanotubes. Microchim Acta 184:4287\u0026ndash;4294\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiao X et al (2022) Lateral flow immunoassay based on time-resolved fluorescence microspheres for rapid and quantitative screening CA199 in human serum. Int J Mol Sci 23:9991\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShin M et al (2024) Highly sensitive multiplexed colorimetric lateral flow immunoassay by plasmon-controlled metal\u0026ndash;silica isoform nanocomposites: PINs. Nano Converg 11:42\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKang H et al (2013) Near-infrared SERS nanoprobes with plasmonic Au/Ag hollow-shell assemblies for in vivo multiplex detection. Adv Funct Mater 23:3719\u0026ndash;3727\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang Y et al (2018) Establishment of time-resolved fluorescence immunochromatographic assay for detection of carbohydrate antigen 19\u0026thinsp;\u0026ndash;\u0026thinsp;9. Sheng Wu Gong Cheng Xue Bao 34:1012\u0026ndash;1018\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu C-W, Tian J-H, Ye J-J, Chang H-W, Tsai Y-C (2021) Construction of a label-free electrochemical immunosensor based on Zn-Co-S/graphene nanocomposites for carbohydrate antigen 19\u0026thinsp;\u0026ndash;\u0026thinsp;9 detection. Nanomaterials 11:1475\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWu S et al (2024) Collard-like Bi2S3@ Au nanocomposites-based label free electrochemical immunosensor for quantitative detection of CA19-9. Talanta 127299\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhao T, Jin B (2024) A label-free electrochemical biosensor based on a bimetallic organic framework for the detection of carbohydrate antigen 19\u0026thinsp;\u0026ndash;\u0026thinsp;9. Anal Methods 16:6173\u0026ndash;6182\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHan X, Lin S, Li Y, Cheng C, Han X (2020) Near-infrared photothermal immunoassay for pancreatic cancer biomarker CA 19\u0026thinsp;\u0026ndash;\u0026thinsp;9 on a digital thermometer. Anal Chim Acta 1098:117\u0026ndash;124\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobin X et al (2011) pROC: an open-source package for R and S\u0026thinsp;+\u0026thinsp;to analyze and compare ROC curves. BMC Bioinform 12:77\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWerner WSM, Glantschnig K, Ambrosch-Draxl C (2009) Optical constants and inelastic electronsScattering data for 17 elemental metals. J Phys Chem Ref Data 38:1013\u0026ndash;1092\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSt\u0026ouml;ber W, Fink A, Bohn E (1968) Controlled growth of monodisperse silica spheres in the micron size range. J Colloid Interface Sci 26:62\u0026ndash;69\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMartin MN, Basham JI, Chando P, Eah S-K (2010) Charged gold nanoparticles in non-polar solvents: 10-Min synthesis and 2D self-assembly. Langmuir 26:7410\u0026ndash;7417\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBock S et al (2021) Lateral flow immunoassay with quantum-dot-embedded silica nanoparticles for prostate-specific antigen detection. Nanomaterials 12:33\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eParolo C et al (2020) Tutorial: Design and fabrication of nanoparticle-based lateral-flow immunoassays. Nat Protoc 15:3788\u0026ndash;3816\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6695327/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6695327/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePancreatic ductal adenocarcinoma (PDAC) is linked to high incidence and mortality rates because it is often detected in later stages, when the prognosis is poor. However, the current state-of-the-art methods for diagnosing early PDAC tend to be invasive, time-consuming, and unreliable, primarily due to the difficulties associated with the early detection of pancreatic cancers. Here, we report a quick and sensitive method for the early diagnosis of PDAC using a signal-enhanced lateral flow immunoassay called SELFI. We developed SELFI, which can generate a strong colorimetric signal through multiple hotspots formed by plasmonic gold nanoparticles (AuNPs) assembled on a silica bead. Our SELFI assay achieved a 28-fold increase in the limit of detection compared to conventional lateral flow immunoassays using 20 nm AuNPs, providing results within 15 min. We demonstrated that SELFI can be utilized for the early diagnosis of PDAC, as indicated by a receiver operating characteristic curve and a larger area under the curve compared to the enzyme-linked immunosorbent assay (****\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). SELFI's effective diagnostic features could enhance the timely identification of PDAC and may also serve in the early diagnosis of a range of other diseases.\u003c/p\u003e","manuscriptTitle":"Early Diagnosis of Pancreatic Ductal Adenocarcinoma by Signal-Enhanced Lateral Flow Immunoassay: SELFI","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-02 06:46:39","doi":"10.21203/rs.3.rs-6695327/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"fc154bba-2493-422c-845e-9c08dc737480","owner":[],"postedDate":"June 2nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":49275245,"name":"Biological sciences/Cancer/Paediatric cancer"},{"id":49275246,"name":"Biological sciences/Biotechnology/Nanobiotechnology/Nanostructures"},{"id":49275247,"name":"Biological sciences/Biotechnology/Nanobiotechnology/Biosensors"}],"tags":[],"updatedAt":"2026-03-26T07:11:29+00:00","versionOfRecord":{"articleIdentity":"rs-6695327","link":"https://doi.org/10.1038/s41467-026-69204-7","journal":{"identity":"nature-communications","isVorOnly":false,"title":"Nature Communications"},"publishedOn":"2026-02-05 05:00:00","publishedOnDateReadable":"February 5th, 2026"},"versionCreatedAt":"2025-06-02 06:46:39","video":"","vorDoi":"10.1038/s41467-026-69204-7","vorDoiUrl":"https://doi.org/10.1038/s41467-026-69204-7","workflowStages":[]},"version":"v1","identity":"rs-6695327","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6695327","identity":"rs-6695327","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}