A machine-learning powered liquid biopsy predicts response to Paclitaxel plus Ramucirumab in advanced gastric cancer: Results from the prospective IVY trial

A machine-learning powered liquid biopsy predicts response to Paclitaxel plus Ramucirumab in advanced gastric cancer: Results from the prospective IVY trial | 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 Research Article A machine-learning powered liquid biopsy predicts response to Paclitaxel plus Ramucirumab in advanced gastric cancer: Results from the prospective IVY trial Katsutoshi Shoda¹², Caiming Xu¹³, Takeshi Nagasaka⁴, Daisuke Ichikawa², and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7059542/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 22 Nov, 2025 Read the published version in Molecular Cancer → Version 1 posted 13 You are reading this latest preprint version Abstract Background Paclitaxel plus ramucirumab (PTX + RAM) is a widely used second-line treatment for advanced gastric cancer, yet no validated biomarkers exist to predict therapeutic response. Identifying non-invasive predictors could enable patient stratification and optimize outcomes. Methods We conducted a prospective observational multicenter study (IVY trial; NCT06490055) enrolling 115 patients with advanced gastric cancer treated with PTX + RAM. Serum was collected prior to the initiation of treatment. Small RNA sequencing identified differentially expressed exosomal microRNAs (exo-miRNAs) between responders and non-responders. Machine learning and logistic regression were employed to construct a predictive model, which was subsequently validated using quantitative real-time polymerase chain reaction (qRT-PCR) in the entire cohort. Results Ten candidate exo-miRNAs were initially discovered, and a five-miRNA panel (miR-10a-5p, miR-25-5p, miR-125a-5p, miR-139-5p, and miR-450a-5p) was selected via stepwise elimination. This 5-exo-miRNA model achieved high accuracy in distinguishing responders from non-responders (AUC = 0.84). When combined with body mass index (BMI), the composite model (EXEMPLAR) demonstrated enhanced predictive performance (AUC = 0.87). High-risk patients exhibited significantly shorter progression-free survival (PFS: median, 1.9 vs. 4.2 months, p = 0.019) and overall survival (OS: median, 1.1 vs. 1.7 years, p < 0.001). Decision curve analysis confirmed the clinical benefit of the model. A nomogram was developed to facilitate personalized risk assessment. Conclusions This study identifies and validates a novel 5-exo-miRNA panel for predicting response to second-line PTX plus RAM therapy in gastric cancer. The combined exosomal signature and BMI risk model provides a clinically applicable, non-invasive tool for personalized treatment selection. ClinicalTrials.gov Identifier: NCT06490055 Gastric cancer Liquid biopsy Exosomal microRNAs Ramucirumab Paclitaxel Machine learning Biomarker Figures Figure 1 Figure 2 Figure 3 Figure 4 INTRODUCTION Gastric cancer is one of the leading causes of cancer-related deaths worldwide [ 1 ]. Combination therapy with fluoropyrimidine and platinum agents is one of the major first-line treatments for unresectable or metastatic gastric cancer [ 2 ]. In addition, the prolonged survival benefits from several other regimens, including paclitaxel and ramucirumab (RAM), have also been demonstrated in various randomized trials [ 3 , 4 ]. Regarding second-line treatments, the recent RAINBOW trial showed that paclitaxel (PTX) plus RAM exhibited significantly better overall survival (OS) compared to PTX plus placebo in patients with previously treated advanced gastric cancer. Consequently, this treatment regimen is now recognized as the standard second-line chemotherapy [ 4 ]. The latest NCCN guidelines also recommend PTX plus RAM as the standard treatment for second-line chemotherapy, with a response rate of approximately 30% and a disease control rate of approximately 70% [ 2 , 4 ]. However, grade 4 neutropenia and febrile neutropenia have been reported in approximately 25% and 5% of patients who receive this treatment, respectively [ 5 ]. In other words, ~ 30% of patients who do not derive any benefit from second-line treatment will not only experience disease progression but will also suffer from the toxicity of such therapy, including a loss of strength and energy. In addition, they may be forced to discontinue or terminate gastric cancer treatment without receiving other treatments that might have provided initial therapeutic efficacy. These patients, who are unable to undergo curative surgery and have transitioned to second-line therapy due to failure of first-line treatment, often face limited survival time. Therefore, selecting the proper treatment in a timely manner is particularly crucial in this setting. The development of clinically useful biomarkers that accurately predict treatment response is essential for guiding therapeutic decisions and improving the clinical management of this population. This highlights the imperative clinical need for the availability of molecular biomarkers that can help predict resistance to second-line therapy and select the subset of gastric cancer patients who will clinically benefit from such a treatment regimen. In the context of markers that can predict response to second-line therapy in gastric cancer, the field remains in its infancy. Ideally, considering the biological effects of chemotherapy, such a treatment prediction should be performed in biospecimens obtained prior to the initiation of any treatment. Previous studies have measured tumor tissue levels of vascular endothelial growth factor receptor (VEGFR) and VEGF-related genes, which did not exhibit any predictive potential despite their association with the mechanism(s) of action of RAM [ 6 – 8 ]. Furthermore, predictive biomarkers based on tissue examination are not as attractive due to their invasiveness and the challenges associated with collecting tumor tissues from metastatic tumors before initiating second-line treatment. In this regard, liquid biopsies based on tumor-derived exosomal cargo are emerging rapidly, offering several distinct advantages [ 9 ]. First, exosomes are extracellular vesicles that facilitate intercellular communication by actively transmitting cellular components, including proteins, DNA, RNA, and various noncoding RNAs [ 10 ]; hence, a significant number of tumor-derived molecular signatures can be readily interrogated in exosomes [ 11 ]. Second, the protective shell of the exosomal membranes mitigates the degradation of molecular entities in bodily fluids, thereby dramatically enhancing the stability of exosomal cargo expression [ 12 ]. These key features, along with the fact that exosome-based liquid biopsies are less invasive and increase patient comfort, are emerging as a paradigm-shifting clinical scenario [ 13 ]. Accordingly, we undertook a systematic and comprehensive study by performing genomewide transcriptomic profiling of exosomal microRNAs (exo-miRNAs) to help identify and develop predictive biomarkers of response to second-line therapy, analyzing pre-treatment blood specimens from a prospective clinical cohort, the IVY-Trial [ 14 ]. We successfully established an exo-miRNA-based liquid biopsy assay that can help predict therapeutic efficacy against PTX plus RAM therapy in gastric cancer patients, allowing for a non-invasive and personalized pre-therapeutic selection approach in these patients. MATERIALS AND METHODS Patient cohorts In this study, we analyzed a cohort of 162 patients enrolled in a prospective clinical trial (the IVY study) [ 14 ]. This longitudinal trial was conducted at multiple institutions in Japan as part of the University Hospital Medical Information Network’s Clinical Trials Registry (registry number UMIN000033376) and ClinicalTrials.gov registration (NCT06490055) to evaluate the efficacy of second-line treatment in patients with gastric cancer. The eligibility and exclusion criteria were reported previously [ 14 ]. Among this cohort, 115 patients with gastric cancer received PTX plus RAM treatment as second-line chemotherapy between October 1, 2018, and October 31, 2021. Blood samples were collected just prior to second-line therapy, processed for serum isolation, and frozen in a -80°C freezer until use. Supplementary Table S1 describes the characteristics of all patients within the clinical cohort. A written informed consent was obtained from all patients, and Institutional Review Board approval was obtained from each participating institution. The study was conducted in compliance with the Declaration of Helsinki. Treatment methods The RAM plus solvent-based (sb)-PTX regimen consisted of RAM (80 mg/m² administered intravenously on days 1 and 15) and sb-PTX (80 mg/m² administered intravenously on days 1, 8, and 15) every 4 weeks. RAM plus nanoparticle albumin-bound (nab)-PTX combination therapy comprised of RAM (80 mg/m2 intravenously on days 1 and 15) and nab-PTX (100 mg/m2 intravenously on days 1, 8, and 15) every 4 weeks [ 2 , 4 , 15 , 16 ]. Exosomal RNA extraction from serum specimens To prepare libraries for small RNA sequencing, exosomes were first isolated from 400 µL of serum using the exoRNeasy Midi Kit (Qiagen, Valencia, CA), followed by RNA extraction using the miRNeasy Kit (Qiagen, Hilden, Germany). For the real-time quantitative reverse-transcription polymerase chain reaction (qRT-PCR) analysis, total exosomal RNA was isolated from 200 µL of serum using the Total Exosome Isolation Kit (Invitrogen, Waltham, MA) and the miRNeasy kit (Qiagen, Valencia, CA), according to the manufacturer’s instructions. Small RNA sequencing and data analysis Exosomal small RNA libraries were prepared using the NEXTflex Small RNA-Seq Kit v3 (PerkinElmer, Waltham, MA, USA). The D1000 TapeStation system (RRID: SCR_018435) was used for quality assessment of the sequencing libraries prior to small RNA sequencing on the Illumina NovaSeq platform (RRID: SCR_016387). Subsequently, the raw fastq sequencing data were generated after paired-end sequencing and evaluated using the FastQC tool (RRID: SCR_014583). Next, adapter trimming and low-quality read filtering were performed using the Cutadapt tool (RRID: SCR_011841). Thereafter, miRNAs were aligned and quantified using the miRDeep2 modules with miRBase (version 22, RRID: SCR_01082). Finally, miRNA abundance for each case was normalized based on counts per million (CPM). Real-time quantitative reverse-transcription polymerase chain reaction assays miRCURY LNA RT Kit (Qiagen) was used for reverse transcription of RNA to complementary DNA (cDNA). SensiFAST SYBR Lo-ROX Kit (Bioline, London, UK) and QuantStudio 6 Flex RT-PCR System (Applied Biosystems, Foster City, CA, USA, RRID: SCR_020239) were used for performing the qRT-PCR assays. The expression level of target exo-miRNAs was normalized using miR-16-5p by the 2-ΔCt method. Study design and evaluation Tumor evaluation by imaging was performed every 2 to 3 months, and treatment response was assessed using the Response Evaluation Criteria in Solid Tumors (RECIST) ver. 1.1 [ 17 ]. Regarding response to second-line therapy, patients who exhibited a partial response (PR) or stable disease (SD) were considered responders. In contrast, those with disease progression (PD) were defined as non-responders. No patient in the study had a complete response to treatment. Progression-free survival (PFS) was defined as the time from enrollment to the first occurrence of progression or death from any cause after initiation of second-line therapy. Overall survival (OS) was defined as the time from enrollment to death or last contact. Statistical analysis Differential expression analysis of miRNAs between responders and non-responders was performed using the R package 'limma' (version 4.2). Differential miRNA expression (up- and down-regulated miRNAs) was presented as volcano plots using the 'ggplot2' package (version 3.3.6), and heatmaps were generated through the 'pheatmap' package (version 1.0.12). Receiver operating characteristic (ROC) curves to evaluate biomarker performance were established using the 'pROC' package (version 1.18.0). Univariate and multivariate logistic regression were performed using the R package 'DAG' (version 1.25.4). The Mann-Whitney U test was used to analyze continuous variables. Survival time analysis included the construction of Kaplan–Meier survival curves for groups based on univariate predictors, which was performed using the “survival” and “survminer” packages in R, with differences assessed via the log-rank test. Multivariate analysis of survival-related clinical factors was performed using the Cox proportional hazards regression model. Nomograms for the Cox hazards regression model were performed using the 'rms' package (version 6.3-0). Statistical significance was determined at a two-tailed P-value threshold of less than 0.05. Statistical analyses were performed using the R statistical software environment (version 4.3.2, RRID: SCR_001905). RESULTS Genome-wide profiling identified 10 circulating exo-miRNA candidates for differentiating patients with and without response to second-line treatment in gastric cancer The primary objective of this study was to identify clinically relevant exo-miRNAs as non-invasive biomarkers for predicting treatment response to PTX plus RAM treatment in gastric cancer. Toward this end, we designed a three-step study that included: i) biomarker discovery by undertaking genome-wide expression profiles of circulating exo-miRNAs that could predict response to PTX plus RAM treatment in gastric cancer, ii) clinical validation for the predictive performance of circulating exo-miRNA biomarker panels, and iii) establishment of a model to predict therapeutic efficacy for routine clinical application (Supplementary Fig. S1 and Supplementary Table S1 ). A biomarker discovery effort based on systematic, comprehensive, and unbiased genome-wide small RNA sequencing was initially undertaken using a total of 28 serum exo-miRNA specimens from responders and non-responders treated with second-line therapy. To gain an understanding of the circulating exo-miRNA profiles in gastric cancer patients immediately prior to second-line treatment, a comparative analysis was conducted between responders and non-responders. Using the criteria of Log2 Foldchange > 1.0 and p-value < 0.05, we identified 19 downregulated and 11 upregulated miRNAs (Fig. 1 A). We next narrowed down our miRNA selection by applying the condition of an average expression level [log2 (CPM + 1)] ≥ 1 and a p-value less than 0.01. Finally, a panel of 10 differentially expressed exo-miRNAs that discriminated between responders and non-responders was identified and prioritized as potential biomarker candidates for further investigations. Six miRNAs (hsa-miR-10a-5p, hsa-miR-10b-5p, hsa-miR-125a-5p, hsa-miR-139-5p, hsa-miR-335-5p, and miR-379-5p) were significantly upregulated, while the remaining four (hsa-miR-25-5p, hsa-miR-450a-5p, hsa-miR-548ad-5p, and hsa-miR-624-5p), were significantly downregulated in non-responder patients (Fig. 1 B). Ridgeline plots visually depict the extent of distinction between responders and non-responders for each candidate biomarker, utilizing their normalized expression level (Z scores). Importantly, logistic regression analysis has demonstrated that all six upregulated miRNAs and 4 downregulated miRNAs were strongly associated with the response to paclitaxel plus ramucirumab as a second-line treatment in GC (Fig. 1 C and Supplementary Table S2 ). The radar chart also visually depicts the absolute coefficients of the logistic regression model for the 10-exo-miRNA panel (Fig. 1 D). In the discovery cohort, the AUC of this 10 exo-miRNA panel was 1.00 (95% CI, 1.00–1.00; 100% sensitivity, 100% specificity, 100% PPV, 100% NPV, Supplementary Fig. S2 ). Taken together, these biomarker discovery results support our initial hypothesis that circulating exo-miRNA markers may provide superior sensitivity and specificity for predicting response to PTX plus RAM treatment in patients with gastric cancer. Potential candidate exo-miRNAs involved in gastric cancer progression To understand the contribution of our candidate circulating exo-miRNAs to gastric cancer progression, we next investigated their biological significance by identifying the specific growth signaling genes downstream targets for these microRNAs. miRNet 2.0 ( https://www.mirnet.ca ) [ 18 ], a comprehensive knowledge base that integrates high-quality miRNA-target interaction data, was used to predict binding sites involving candidate miRNAs and information on transcription factors affecting target genes. It was intriguing to note that the target gene clusters for the identified candidate miRNAs revealed a significant involvement not only in the MAPK and Wnt signaling pathways, which are oncogenic pathways involved in gastric cancer progression, but also in the VEGF pathway, the primary mechanistic signaling pathway responsible for RAM (Fig. 1 E). Furthermore, the high-risk group, stratified by candidate exo-miRNAs in the discovery cohort, exhibited a significantly poorer prognosis after initiation of second-line therapy compared to the low-risk group (Supplementary Fig. S3 ). These results indicate that the candidate exo-miRNAs are not only predictive biomarkers for sensitivity to PTX plus RAM treatment but also potential biomarkers for gastric cancer progression and prognosis after second-line treatment. Performance of the exo-miRNA panel for discriminating patients with response to second-line PTX plus RAM treatment in gastric cancer To establish a routine clinical predictive assay based on the discovered miRNA biomarkers, qRT-PCR assays were performed on blood specimens obtained from patients in a prospective clinical trial. Information on clinicopathologic factors for this prospective clinical cohort is shown in Supplemental Table S1 . The expression levels of these 10 candidate exo-miRNAs were examined in a cohort of 115 patients, including 80 responders and 35 non-responders. The results from the PCR assays on the serum specimens were consistent with those in our discovery cohort, indicating that all 10 exo-miRNAs were readily detectable and confirming the robustness of our biomarker discovery. Subsequently, these candidate exo-miRNAs were incorporated into a logistic regression model to train a risk prediction algorithm for identifying the response to PTX plus RAM treatment in patients with gastric cancer. For each of the exo-miRNAs in this panel development, coefficients and constants derived from the logistic regression equation were applied as follows to calculate a risk score for its ability to predict a lack of response to PTX plus RAM therapy: 10-exo-miRNA panel score = (-6.56011 x exo-miR10a) + (-10.48902 x exo-miR10b) + (-205.95086 x exo-miR125a-5p) + (20.62076 x exo-miR139-5p) + (0.32001 x exo-miR25-5p) + (-1.00614 x exo-miR335-3p) + (-0.0024514 x exo-miR379-5p) + (0.76029 x exo-miR450a-5p) + 0.023331 x exo-miR548ad) + (-0.097031 x exo-miR624-5p) − 0.61411. The combined analysis of these markers with the 10-exo-miRNA panel revealed that this signature was significantly superior in overall predictive accuracy, with the risk score being significantly higher in non-responders compared to responders (p < 0.001, Supplementary Fig. S4A), and a robust corresponding AUC value of 0.83 (Supplementary Fig. S4B). The overall predictive accuracy of this composite panel is consistent with the potential obtained from the discovery cohort study. The prognostic potential of the exo-miRNA panel in predicting survival outcomes in patients with second-line treatment in gastric cancer To evaluate the prognostic potential of our miRNA panel, we performed a survival analysis by comparing overall survival (OS) and progression-free survival (PFS) in patients categorized as either high- or low-risk based on the 10-exo-miRNA panel. The median follow-up duration for the entire cohort was 18.9 months (95% CI = 16.7–21.4). It was interesting to observe that, in this prospective cohort, high-risk patients exhibited a significantly worse prognosis compared to low-risk patients (PFS: p = 0.019; OS: p < 0.001; Supplementary Fig. S5A, B). These results indicate that, in addition to its ability to predict therapeutic response in gastric cancer patients undergoing second-line therapy, this exo-miRNA panel offers significant prognostic potential in this disease. Establishment of a clinically attractive assay for non-invasive identification of therapeutic response prediction to second-line treatment in gastric cancer Simpler analysis methods are needed for clinical applications, and reducing the number of markers to be analyzed is economically viable. To develop a clinically feasible and cost-effective assay that includes only the minimal number of markers necessary to maintain the overall predictive performance of the exo-miRNAs, a systematic stepwise backward elimination method was used to prioritize the most pertinent biomarker candidates. This statistical strategy resulted in a reduced panel of five exo-miRNA candidates: hsa-miR-10a-5p, hsa-miR-25-5p, hsa-miR-125a-5p, hsa-miR-139-5p, and hsa-miR-450a-5p. The reduced panel of candidate miRNAs was then used to develop a logistic regression equation to recalibrate the final risk prediction model in the clinical cohort as follows: 5-exo-miRNA panel score = (21.31891 x exo-miR10a) + (0.33244 x exo-miR25-5p) + (-315.55693 x exo-miR125a-5p) + (23.60580 x exo-miR139-5p) + (0.60430 x exo-miR450a-5p) − 0.60185. This recalibrated model was once again applied to the entire clinical cohort which revealed an unchanged predictive performance, with risk scores significantly higher in the higher in the group with a lack of response vs. the response group (Fig. 2 A). Notably, the predictive AUC value of this 5-exo-miRNA panel was 0.84, which was higher than the predictive performance of the larger pool of 10 markers (Fig. 2 B). Furthermore, the performance of the exo-miRNA panel with this reduced set of markers revealed a higher predictive performance improvement of 88% specificity, indicating the ability to identify gastric cancer patients who will benefit from PTX plus RAM therapy with very high performance (Fig. 2 C). Taken together, these results are very promising and highlight that the reduced exo-miRNA panel is a highly robust, clinically attractive, and inexpensive assay for predicting treatment response in gastric cancer patients undergoing PTX plus RAM therapy. The prognostic potential of the 5-exo-miRNA panel in gastric cancer patients undergoing second-line treatment To evaluate the prognostic potential of our recalibrated panel, we performed a survival analysis by comparing OS and PFS in patients categorized as either high or low-risk based on the 5-exo-miRNA panel. High-risk patients had a significantly poorer prognosis than low-risk patients (PFS: p = 0.019; OS: p < 0.001; Fig. 2 D, Supplementary Fig. S6). It is worth noting that the median progression-free survival (PFS) for high-risk patients is 1.9 months. In comparison, low-risk patients have a significantly different median PFS of 4.2 months (Fig. 2 D). Furthermore, multivariate Cox proportional hazards regression analysis revealed that our 5-exo-miRNA panel was an independent prognostic factor along with body mass index (BMI) in this prospective cohort (HR = 2.34, 95% CI = 1.56–4.81, p < 0.01; Table 1 ). These results reemphasize that this exo-miRNA panel is a powerful prognostic factor as well as has the ability to predict treatment response in gastric cancer patients undergoing second-line therapy. Table 1 Cox proportional hazards regression analysis of overall survival in the IVY study Univariate Multivariate OR (95% CI) P value OR (95% CI) P value Age, ≥ 65 vs. < 65 years 0.98 (0.62–1.55) 0.93 Gender, Male vs. Female 0.77 (0.50–1.20) 0.24 Body mass index, ≥ 18 vs. < 18 0.50 (0.30–0.83) < 0.01 0.55 (0.33–0.91) 0.02 Primary tumor location, Upper third vs. Lower two-thirds 0.65 (0.41–1.03) 0.07 Histological type, Diffuse type vs. Intestinal type 1.42 (0.92–2.19) 0.12 Metastatic site, Hematogenous vs. Peritoneal or Lymphatic 1.09 (0.68–1.74) 0.72 History of primary lesion resection, Present vs. Absent 0.99 (0.63–1.54) 0.95 HER2 status, Positive vs. Negative 0.63 (0.36–1.12) 0.12 5-exo-miRNA panel risk score, High vs. Low 2.46 (1.56–3.86) < 0.01 2.34 (1.48–3.69) < 0.01 OR, odds ratio, CI, confidence interval. A combination of exosome-based miRNA panel and BMI significantly improved the predictive accuracy of response to second-line therapy in gastric cancer patients Considering the current landscape of clinical risk factors involved in the progression of gastric cancer patients with distant metastases, we investigated whether a risk stratification model combining currently used clinicopathological risk factors (i.e., BMI, HER2 status, gender, metastatic site, tumor location of the primary tumor, age, and history of primary tumor resection) with our exo-miRNA panel could further improve predictive accuracy of treatment response in gastric cancer patients receiving second-line therapy. Univariate analysis revealed that BMI (BMI < 18; p = 0.002) and the 5-exo-miRNA panel (p < 0.001) were significantly associated with a lack of treatment response in gastric cancer patients receiving PTX plus RAM therapy (Fig. 3 A). Subsequent multivariate analysis including only significant variables from the univariate model revealed that BMI (BMI < 18; p = 0.032) and exo-miRNA panel (p < 0.001) emerged as independent risk factors for a lack of response to PTX plus RAM treatment. (Fig. 3 A). We next examined whether the robustness of our 5-exo-miRNA panel could be further improved by including BMI, which has been independently found to be associated with a lack of treatment response in multivariate analysis. In support of our hypothesis, we found that combining the exo-miRNA panel with BMI improved the AUC from 0.84 to 0.87, thus significantly improving the robustness of this clinical cohort in predicting a lack of response to treatment. (Risk model AUC, 0.87; 95% CI, 0.80–0.93) (Fig. 3 B). We defined this composite signature score as EXEMPLAR (EXo-miRNAs-based Model for Predicting Response to PacLitAxel plus Ramucirumab Treatment). EXEMPLAR was significantly superior in overall predictive accuracy, with risk scores significantly higher in non-responders vis-à-vis responders (p < 0.001, Fig. 3 C). The detailed distribution of EXEMPLAR risk scores in gastric cancer patients is shown in the waterfall plot (Fig. 3 D). The waterfall plot analysis also depicts that EXEMPLAR successfully discriminated between responders and non-responders with a sensitivity of 86%, specificity of 82%, and a true negative of 65/70 (92.9%). These findings again highlight that, although this exosome-based miRNA assay was quite robust on its own, combining it with BMI results in a significant improvement in overall predictive accuracy, emphasizing its potential for translation into the clinic for predicting patients who lack a response to PTX plus RAM treatment. An exosome-based liquid biopsy assay offers significant advantages for predicting therapeutic outcomes in patients with gastric cancer undergoing second-line treatment. In current clinical practice, there are no standardized tests or biomarkers that can predict the efficacy of PTX plus RAM therapy for gastric cancer patients prior to treatment. Therefore, the standard second-line chemotherapy, RAM plus PTX, is recommended for all patients who have failed to respond to first-line therapy or who are unable to continue treatment due to various side effects. Although PTX plus RAM is undoubtedly a proper regimen that has shown some efficacy in clinical trials, the cytotoxic nature of anticancer therapy may not be helpful for the 30% of patients who receive treatment but will not benefit from it. Therefore, the clinical utility of a treatment strategy should be evaluated by weighing the benefits against the harms and therapeutic effects. To further investigate the clinical significance of our liquid biopsy assay, we performed a decision curve analysis. As shown in Fig. 3 D, the x-axis represents the threshold probability for a lack of response to PTX plus RAM treatment, and the y-axis represents the net benefit achieved. The decision curve analysis revealed that EXEMPLAR achieved a higher net benefit score across most threshold probabilities compared to PTX plus RAM treatment for all patients, and no patients received PTX plus RAM treatment (Fig. 3 E). For example, at a threshold probability of 0.30, liquid biopsy with EXEMPLAR achieved a significantly higher net benefit of about 0.18 for PTX plus RAM treatment. In contrast, the net benefit based on the intervention strategy for all patients was significantly lower, at almost zero. These findings suggest that, in terms of disadvantage avoidance, EXEMPLAR provides significantly higher clinical benefits compared to interventions in all patients or no intervention in any patients. Establishment of an EXEMPLAR-based risk nomogram in gastric cancer patients undergoing second-line treatment for a personalized therapeutic approach Subsequently, we evaluated the impact of the EXEMPLAR, which was established using the statistically significant factors exo-miRNA panel and BMI, on the prognostic value in patients undergoing second-line treatment for gastric cancer. EXEMPLAR-high patients had a significantly poorer prognosis in both OS and PFS compared to low-risk patients (OS: p < 0.001; PFS: p < 0.001; Fig. 4 A, B). The details of PFS and OS for each case are shown in a swimmer plot in Fig. 4 C, and the survival analysis for EXEMPLAR-low and EXEMPLAR-high patients is summarized in Table 2 . It should be noted that the median PFS for EXEMPLAR-low and EXEMPLAR-high patients was 4.8 and 1.9 months, respectively, a difference of more than 2-fold. Regarding OS, it is also noteworthy that EXEMPLAR-high patients had no 3-year survivors after 2nd-line initiation of therapy, whereas EXEMPLAR-low patients achieved a 23% 3-year survival rate. Table 2 Survival according to EXEMPLAR status Overall EXEMPLAR high EXEMPLAR low P value Progression-free survival 3-month PFS rate, % (95% CI) 60.0% (50.9–68.4%) 28.9% (17.7–43.4%) 80.0% (69.2–87.7%) < 0.01 6-month PFS rate, % (95% CI) 25.2% (18.2–33.9) 11.1% (4.8–23.4%) 34.3% (24.3– 46.0%) < 0.01 Median PFS, months (95% CI) 3.97 (3.00–4.67) 1.93 (1.27–2.13) 4.81 (4.10–5.80) < 0.01 Hazard Ratio of Progression (95% CI) - 2.12 (1.44–3.12) Reference Overall Survival 1-year OS rate, % (95% CI) 67.8% (58.8–75.7%) 55.6% (41.2–69.1%) 75.7% (64.5–84.2%) < 0.01 2-year OS rate, % (95% CI) 22.6% (15.9–31.1%) 8.9% (3.5–20.7%) 31.4% (21.8–43.0%) < 0.01 Median OS, months (95% CI) 1.4 (1.2–1.5) 1.1 (0.9–1.3) 1.7 (1.4–1.9) < 0.01 Hazard Ratio of Death (95% CI) - 2.56 (1.67–3.92) Reference * P-values refer to comparisons between the Exo-miRs model high vs. low, CI, confidence interval. Finally, to facilitate the clinical translation of our findings, we developed a prognostic risk nomogram for PFS. These results indicated that our exo-miRNA panel score carried a higher weight than BMI, once again highlighting its clinical significance for an expedient translation into clinical practice (Fig. 4 D). Taken together, these data suggest that combining our exo-miRNA panel with the nutritional indicator BMI could provide a more sophisticated personalized medicine approach by proposing a prognostic risk probability model for GC patients undergoing PTX plus RAM treatment. DISCUSSION With today's advances in cancer treatment and the increasing availability of various treatment options, it is essential to develop treatment strategies that provide optimal care for each patient [ 19 ]. Many drugs have been developed for the treatment of patients with advanced gastric cancer, and their efficacy has been demonstrated in clinical trials [ 3 , 6 , 15 ]. We have entered an era in which long-term survival is possible even for patients with advanced gastric cancer if they respond to chemotherapy [ 20 ]. However, the prognosis is still poor for patients with advanced gastric cancer who are untreated or do not respond to treatment [ 21 ]. Therefore, the importance of selecting second-line treatment is evident, especially for patients who have failed to respond to first-line therapy, and establishing a predictive risk model for treatment efficacy is undoubtedly desirable. Our study is a step in this direction, and we have successfully identified a novel exo-miRNAs model that robustly identifies treatment response in gastric cancer patients undergoing second-line therapy through systemic and comprehensive biomarker discovery and validation. More interestingly, our established exo-microRNA model was also able to individually stratify the prognosis of gastric cancer patients undergoing treatment after initiation of therapy. These results highlight the potential clinical significance of a novel exo-miRNAs model for predicting treatment response in gastric cancer patients undergoing second-line therapy. To date, there are no established predictive biomarkers for PTA plus RAM treatment, the main second-line treatment for gastric cancer. For patients with gastric cancer, the NCCN guidelines recommend the use of fluoropyrimidine and platinum-based chemotherapy as one of the main first-line therapies, with second-line therapy for patients who do not respond to first-line therapy or who have difficulty continuing treatment. PTX plus RAM therapy is the primary second-line treatment, based on the results of the Phase III RAINBOW and REGARD trials [ 3 , 4 ]. Other treatment options include several immune checkpoint inhibitors, irinotecan, trifluridine, and other molecularly targeted agents such as trastuzumab deruxtecan and zolbetuximab [ 21 – 25 ]. However, persistent lack of response to treatment may, in some cases, necessitate termination of treatment during the first or second line of treatment. For patients who fail to respond to first-line treatment, failure to respond to second-line treatment also results in a loss of strength and energy, forcing patients to discontinue or terminate treatment without receiving other treatments that might have been effective. These realities in clinical practice highlight the need to develop tests and markers for predicting the effectiveness of second-line treatment. Therefore, several previous studies have attempted to find biomarkers to predict the therapeutic efficacy of PTX plus RAM therapy [ 5 ]. RAINBOW investigators attempted to develop a biomarker targeting VEGF, the primary mechanism of action of RAM, and examined the association between plasma levels of VEGF-D and therapeutic response, but were unable to demonstrate a significant therapeutic predictive marker effect of ramucirumab in gastric cancer [ 5 , 26 , 27 ]. Lorenzo et al. suggested that an early increase in soluble VEGFR-2 after one cycle of treatment may be a potential predictive biomarker, and Raghav et al. developed a gene signature using a random forest machine learning model with resected tissue for predicting PTX efficacy [ 26 , 27 ]. Ideally, however, the patient's condition should be assessed in real-time immediately prior to the start of treatment, given the anticipated impact of primary treatment on the outcome of second-line treatment. In this regard, it is worth noting that our study is prospective, in which blood was collected immediately before treatment for a specific regimen. To our knowledge, this is the first report on a biomarker that predicts treatment response for PTX plus RAM using liquid biopsies from a prospective study conducted just prior to treatment. We employed a systematic and comprehensive biomarker discovery approach to develop a blood-based exo-miRNA panel with superior performance in predicting treatment outcomes in second-line gastric cancer therapy. Interestingly, in a public database analysis, these exo-miRNA biomarkers were shown to be associated with the VEGF pathway, the primary mechanism of action of RAM, supporting our finding that RAM-based therapeutic regimens correlate with therapeutic efficacy [ 8 ]. Furthermore, our exo-miRNA panel demonstrated high predictive ability in a prospective clinical cohort, clearly stratifying patient prognosis after treatment. In addition, other clinical factors correlated with treatment response were analyzed using multivariate logistic regression analysis, and BMI was selected in conjunction with our exo-miRNA panel. In gastric cancer, high pre-treatment BMI and muscle mass are significantly associated with better prognosis after treatment, and nutritional status during chemotherapy has also been reported to correlate with oncologic prognosis28. Conversely, patients with sarcopenia have an increased incidence of treatment-induced side effects, which are also thought to be associated with poor treatment adherence and prognosis [ 29 ]. These phenomena are considered one of the characteristics of gastric cancer treatment and are called BMI paradoxes [ 30 ]. An interesting aspect of this study is that a more robust predictive model was established by combining the exo-miRNA panel, which was explored in relation to tumor molecular biology, with BMI, a patient's nutritional index. From a clinical perspective, our exo-miRNAs model has the potential to enable several important improvements when introduced into clinical practice. (i) The possibility of selecting a treatment alternative to PTX plus RAM for patients expected to be resistant to treatment can be considered. (ii) For patients who are predicted to respond well to treatment, patients and physicians can make treatment choices based on a solid perception of the benefits of treatment, even if some risk of chemotherapy side effects is expected. (iii) For patients receiving first-line therapy who are expected to respond well to second-line therapy, early conversion to second-line therapy may be recommended, given the irreversible side effects of primary therapy. Furthermore, our liquid biopsy-based exo-miRNA model is non-invasive and repeatable, making it potentially adaptable for assessing risk status at multiple time points throughout a patient's clinical course. Its performance in this regard should be examined in future evaluations of multiple liquid biopsies in the same patient. We would like to acknowledge a few potential limitations of our present study. First, although our study was a prospective clinical study focusing on blood immediately before second-line treatment of gastric cancer, prospective studies using larger patient cohorts are also needed before these biomarkers can be examined in actual clinical practice. Second, real-time evaluation of exosomes may enable the monitoring of cancer progression, and evaluating pre- and post-treatment exosome expression in the same patient may be promising. Nevertheless, our current study provides promising evidence for the clinical significance of the exo-miRNA model in predicting treatment response in gastric cancer patients undergoing second-line treatment, marking an important step toward the application of robust molecular biomarkers for risk assessment and management of lethal malignancies, such as advanced metastatic gastric cancer. In conclusion, we have successfully validated noninvasive liquid biopsies in a prospective clinical cohort using genome-wide exo-miRNAs expression profiling to predict treatment response in gastric cancer patients receiving second-line therapy. This is clinically crucial for patients with advanced gastric cancer to develop an optimized treatment plan. Furthermore, our assay provides a framework for a personalized medicine approach, enabling risk-based management and treatment tailored to each patient prior to the initiation of therapy. Declarations Ethics approval and consent to participate : This study was conducted in accordance with the Declaration of Helsinki. All procedures were approved by the Ethics Committee of Kawasaki Medical School. Written informed consent was obtained from all participants prior to enrollment in the study. Consent for publication: Not applicable. Availability of data and materials: The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request. Competing interests: The authors declare that they have no competing interests. Funding: This work was supported by grants CA72851, CA181572, CA187956, CA227602, CA214254, and CA271443 from the National Cancer Institute, NIH. Authors' contributions: Study concept and design: KS, YA, AG. Provision of samples and clinical data: KS, TN, AG. Bioinformatic analysis: CX. Data collection and analysis: KS, CX, YA, AM. Drafting of the manuscript: KS, CX, YA, TN, DI, AG. All authors read and approved the final manuscript. Acknowledgements: The authors thank Drs. Yoh Asahi and Alessandro Mannucci for their thoughtful discussions and advice during this project. References Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-249. Ajani JA, D'Amico TA, Bentrem DJ, et al. Gastric Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2022;20:167-192. Fuchs CS, Tomasek J, Yong CJ, et al. Ramucirumab monotherapy for previously treated advanced gastric or gastro-oesophageal junction adenocarcinoma (REGARD): an international, randomised, multicentre, placebo-controlled, phase 3 trial. Lancet 2014;383:31-39. Wilke H, Muro K, Van Cutsem E, et al. Ramucirumab plus paclitaxel versus placebo plus paclitaxel in patients with previously treated advanced gastric or gastro-oesophageal junction adenocarcinoma (RAINBOW): a double-masked, randomised phase 3 trial. Lancet Oncol 2014;15:1224-35. Xu RH, Zhang Y, Pan H, et al. Efficacy and safety of weekly paclitaxel with or without ramucirumab as second-line therapy for the treatment of advanced gastric or gastroesophageal junction adenocarcinoma (RAINBOW-Asia): a randomised, multicentre, double-masked, phase 3 trial. Lancet Gastroenterol Hepatol 2021;6:1015-1024. Van Cutsem E, Muro K, Cunningham D, et al. Biomarker analyses of second-line ramucirumab in patients with advanced gastric cancer from RAINBOW, a global, randomized, double-blind, phase 3 study. Eur J Cancer 2020;127:150-157. Natsume M, Shimura T, Iwasaki H, et al. Placental growth factor is a predictive biomarker for ramucirumab treatment in advanced gastric cancer. Cancer Chemother Pharmacol 2019;83:1037-1046. Aprile G, Ongaro E, Del Re M, et al. Angiogenic inhibitors in gastric cancers and gastroesophageal junction carcinomas: A critical insight. Crit Rev Oncol Hematol 2015;95:165-78. Shigeyasu K, Toden S, Zumwalt TJ, et al. Emerging Role of MicroRNAs as Liquid Biopsy Biomarkers in Gastrointestinal Cancers. Clin Cancer Res 2017;23:2391-2399. Yu W, Hurley J, Roberts D, et al. Exosome-based liquid biopsies in cancer: opportunities and challenges. Ann Oncol 2021;32:466-477. Tang XH, Guo T, Gao XY, et al. Exosome-derived noncoding RNAs in gastric cancer: functions and clinical applications. Mol Cancer 2021;20:99. Huang T, Song C, Zheng L, et al. The roles of extracellular vesicles in gastric cancer development, microenvironment, anti-cancer drug resistance, and therapy. Mol Cancer 2019;18:62. Kalluri R, LeBleu VS. The biology, function, and biomedical applications of exosomes. Science 2020;367. Tanioka H, Nagasaka T, Uno F, et al. The relationship between peripheral neuropathy and efficacy in second-line chemotherapy for unresectable advanced gastric cancer: a prospective observational multicenter study protocol (IVY). BMC Cancer 2019;19:941. Shitara K, Takashima A, Fujitani K, et al. Nab-paclitaxel versus solvent-based paclitaxel in patients with previously treated advanced gastric cancer (ABSOLUTE): an open-label, randomised, non-inferiority, phase 3 trial. Lancet Gastroenterol Hepatol 2017;2:277-287. Bando H, Shimodaira H, Fujitani K, et al. A phase II study of nab-paclitaxel in combination with ramucirumab in patients with previously treated advanced gastric cancer. Eur J Cancer 2018;91:86-91. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228-47. Fan Y, Siklenka K, Arora SK, et al. miRNet - dissecting miRNA-target interactions and functional associations through network-based visual analysis. Nucleic Acids Res 2016;44:W135-41. Joshi SS, Badgwell BD. Current treatment and recent progress in gastric cancer. CA Cancer J Clin 2021;71:264-279. Alsina M, Arrazubi V, Diez M, et al. Current developments in gastric cancer: from molecular profiling to treatment strategy. Nat Rev Gastroenterol Hepatol 2023;20:155-170. Kang YK, Chen LT, Ryu MH, et al. Nivolumab plus chemotherapy versus placebo plus chemotherapy in patients with HER2-negative, untreated, unresectable advanced or recurrent gastric or gastro-oesophageal junction cancer (ATTRACTION-4): a randomised, multicentre, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2022;23:234-247. Shitara K, Van Cutsem E, Bang YJ, et al. Efficacy and Safety of Pembrolizumab or Pembrolizumab Plus Chemotherapy vs Chemotherapy Alone for Patients With First-line, Advanced Gastric Cancer: The KEYNOTE-062 Phase 3 Randomized Clinical Trial. JAMA Oncol 2020;6:1571-1580. Hironaka S, Ueda S, Yasui H, et al. Randomized, open-label, phase III study comparing irinotecan with paclitaxel in patients with advanced gastric cancer without severe peritoneal metastasis after failure of prior combination chemotherapy using fluoropyrimidine plus platinum: WJOG 4007 trial. J Clin Oncol 2013;31:4438-44. Shitara K, Doi T, Dvorkin M, et al. Trifluridine/tipiracil versus placebo in patients with heavily pretreated metastatic gastric cancer (TAGS): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2018;19:1437-1448. Shitara K, Bang YJ, Iwasa S, et al. Trastuzumab Deruxtecan in Previously Treated HER2-Positive Gastric Cancer. N Engl J Med 2020;382:2419-2430. Fornaro L, Musettini G, Orlandi P, et al. Early increase of plasma soluble VEGFR-2 is associated with clinical benefit from second-line treatment of paclitaxel and ramucirumab in advanced gastric cancer. Am J Cancer Res 2022;12:3347-3356. Sundar R, Barr Kumarakulasinghe N, Huak Chan Y, et al. Machine-learning model-derived gene signature predictive of paclitaxel survival benefit in gastric cancer: results from the randomised phase III SAMIT trial. Gut 2022;71:676-685. Meng Q, Tan S, Jiang Y, et al. Post-discharge oral nutritional supplements with dietary advice in patients at nutritional risk after surgery for gastric cancer: A randomized clinical trial. Clin Nutr 2021;40:40-46. Kawamura T, Makuuchi R, Tokunaga M, et al. Long-Term Outcomes of Gastric Cancer Patients with Preoperative Sarcopenia. Ann Surg Oncol 2018;25:1625-1632. Kamiya H, Komatsu S, Nishibeppu K, et al. Obesity paradox as a new insight from postoperative complications in gastric cancer. Sci Rep 2023;13:10116. Additional Declarations No competing interests reported. Supplementary Files Shodaetal.IVYTrialMSSupplementaryFigures.pdf Shodaetal.IVYTrialMSSupplementaryTables.docx Shodaetal.IVYTrialMSClinicalTrialProtocol.docx Cite Share Download PDF Status: Published Journal Publication published 22 Nov, 2025 Read the published version in Molecular Cancer → Version 1 posted Editorial decision: Revision requested 20 Aug, 2025 Reviews received at journal 17 Aug, 2025 Reviews received at journal 06 Aug, 2025 Reviewers agreed at journal 30 Jul, 2025 Reviewers agreed at journal 29 Jul, 2025 Reviews received at journal 29 Jul, 2025 Reviewers agreed at journal 29 Jul, 2025 Reviewers agreed at journal 28 Jul, 2025 Reviewers agreed at journal 17 Jul, 2025 Reviewers invited by journal 08 Jul, 2025 Editor assigned by journal 07 Jul, 2025 Submission checks completed at journal 07 Jul, 2025 First submitted to journal 06 Jul, 2025 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-7059542","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":482701953,"identity":"c774d593-11e8-44e4-95a6-e629f52ad887","order_by":0,"name":"Katsutoshi Shoda¹²","email":"","orcid":"","institution":"Beckman Research Institute of City of Hope","correspondingAuthor":false,"prefix":"","firstName":"Katsutoshi","middleName":"","lastName":"Shoda¹²","suffix":""},{"id":482701954,"identity":"c3ef5790-9977-4d6e-a005-cac744299748","order_by":1,"name":"Caiming Xu¹³","email":"","orcid":"","institution":"Beckman Research Institute of City of Hope","correspondingAuthor":false,"prefix":"","firstName":"Caiming","middleName":"","lastName":"Xu¹³","suffix":""},{"id":482701955,"identity":"03d870d7-a079-4d42-ac94-1019dcf48c41","order_by":2,"name":"Takeshi Nagasaka⁴","email":"","orcid":"","institution":"Kawasaki Medical School","correspondingAuthor":false,"prefix":"","firstName":"Takeshi","middleName":"","lastName":"Nagasaka⁴","suffix":""},{"id":482701956,"identity":"49ca7830-6dc6-4942-84e7-0e3e693678dd","order_by":3,"name":"Daisuke Ichikawa²","email":"","orcid":"","institution":"University of Yamanashi","correspondingAuthor":false,"prefix":"","firstName":"Daisuke","middleName":"","lastName":"Ichikawa²","suffix":""},{"id":482701957,"identity":"b827e123-2ef7-4492-bde0-8d3a115d921e","order_by":4,"name":"Ajay Goel¹⁵","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAtklEQVRIiWNgGAWjYFCCA0BcYMPAIEGaFoM0krSAgMFhErTwNx6/+OCDwXl5fukGxscVv4jQInHgTLHhDIPbhjPnHGA2PNtHjDUHzqRJ8xjcTjC4kcAm2dhDhA55kJY/BudI0GJw4PgxaSAJ0dLwgwgthgfOMBv2GCQbzpyR2GzY2ECEFrkbxx8++FFhJ88vkXzwYcMfIrQwSJwxgLIYGxgY24jRwt/+AIlHlC2jYBSMglEw0gAAKsU71rX+LscAAAAASUVORK5CYII=","orcid":"","institution":"Beckman Research Institute of City of Hope","correspondingAuthor":true,"prefix":"","firstName":"Ajay","middleName":"","lastName":"Goel¹⁵","suffix":""}],"badges":[],"createdAt":"2025-07-06 19:08:02","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7059542/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7059542/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1186/s12943-025-02526-0","type":"published","date":"2025-11-22T15:56:59+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":86655279,"identity":"be970947-e3a7-47bb-871a-0640edc1f696","added_by":"auto","created_at":"2025-07-14 10:11:47","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":631147,"visible":true,"origin":"","legend":"\u003cp\u003eGenome-wide discovery of circulating exosomal miRNA candidates for predicting response to RAM plus PTX chemotherapy in gastric cancer patients in the small RNA sequencing assay. (A) Volcano plot of differentially expressed miRNAs. The red and blue dots represent up-regulated and down-regulated exosomal miRNAs (exo-miRNAs) in non-responder patients compared to responder patients. (B) A heatmap based on a 10-exo-miRNA panel to predict response to RAM plus PTX chemotherapy. The differences in biomarker expression levels between responders and non-responders are shown in the bar graphs. (C) Ridge-line plots depict the normalized distribution of the expression levels of exo-miRNAs in responders vs. non-responders (left). Forest plot depicting odds ratios and 95% CIs for candidate exo-miRNAs to identify patients with a lack of response to RAM plus PTX treatment (right). (D) Radar chart depicting the absolute coefficients of the logistic regression model for the 10-exo-miRNA panel. (E) Candidate miRNAs-related pathways in the miRnet database. Responder, stable disease or partial response; Non-responder, progressive disease; PTX, paclitaxel; RAM, ramucirumab; OR, odds ratio; CI, confidence interval.\u003c/p\u003e","description":"","filename":"fig1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/2e96aceb9a187e466b5f7cc7.jpg"},{"id":86655266,"identity":"36acc47e-4cb7-485c-a4f6-e9fe318a8d5a","added_by":"auto","created_at":"2025-07-14 10:11:47","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":383991,"visible":true,"origin":"","legend":"\u003cp\u003eClinical validation for the stepwise 5 biomarkers to differentiate between responders and non-responders. (A) Rain-cloud plots with overlapping box and whisker plots illustrate the stepwise 5-exo-miRNA panel scores in patients with a response and those without a response. (B) ROC curve analysis to identify responders and non-responders with an AUC of 0.84 (95% CI 0.76 – 0.92). (C) The confusion matrix was used to identify responders and non-responders with a sensitivity and specificity of 0.66 and 0.88, respectively. (D) Prognostic potential of the candidate 5-exo-miRNA biomarker panel in gastric cancer patients undergoing RAM plus PTX therapy. Kaplan-Meier curves of the progression-free survival for patients with high or low-risk scores in the 5-exo-miRNA panel. ROC curves and Kaplan-Meier curves are shown with 95% CI. PFS, progression-free survival; OS, overall survival; ROC, receiver operating characteristics; AUC, area under the curve; CI, confidence interval.\u003c/p\u003e","description":"","filename":"fig2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/eb4397c2d5c5ece4e02c5f51.jpg"},{"id":86655252,"identity":"9fa160be-dab4-426f-a25f-3a09860498e8","added_by":"auto","created_at":"2025-07-14 10:11:47","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":437389,"visible":true,"origin":"","legend":"\u003cp\u003ePredictive potential of established combination models for response to RAM + PTX therapy in gastric cancer patients. (A) Forest plot depicting odds ratios and 95% confidence intervals for univariate and multivariate clinicopathological variables and a 5-exo-miRNA panel to identify patients with a lack of response to RAM plus PTX treatment. (B) ROC for predicting the response to RAM plus PTX treatment. The green and blue lines represent the BMI and risk model, respectively, constructed from the combination of the 5-exo-miRNA panel and BMI. (C) Violin plots for the EXEMPLAR risk scores in patients with response and a lack of response. (D) The distribution of risk score for EXEMPLAR status. (E) Decision curve analysis to evaluate the performance of the EXEMPLAR. ROC curves are shown with 95% CI. PTX, paclitaxel; RAM, ramucirumab; ROC, receiver operating characteristics; AUC, area under the curve; OR, odds ratio; CI, confidence interval; BMI, Body mass index; EXEMPLAR, exo-miRNAs model to predict treatment response for paclitaxel and ramucirumab.\u003c/p\u003e","description":"","filename":"fig3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/b7b47be72f83d3265c696871.jpg"},{"id":86655255,"identity":"de118175-e233-4235-8f62-8dc80bfaa917","added_by":"auto","created_at":"2025-07-14 10:11:47","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":507781,"visible":true,"origin":"","legend":"\u003cp\u003ePrognostic potential of the established combination model to predict the response to RAM plus PTX treatment in patients with gastric cancer. (A) Kaplan-Meier analysis of the EXEMPLAR for progression-free survival. (B) Kaplan-Meier analysis of the EXEMPLAR for overall survival. (C) Swimmers plot demonstrating the PFS (horizontal boxes) and OS (horizontal purple lines) in EXEMPLAR-high (red) and EXEMPLAR-low (green) gastric cancer patients. (D) The nomogram was derived from the combination of the 5-exo-miRNA panel and BMI. ROC curves and Kaplan-Meier curves are shown with 95% CI. PFS, progression-free survival; OS, overall survival; OR, odds ratio; ROC, receiver operating characteristics; AUC, area under the curve; CI, confidence interval.\u003c/p\u003e","description":"","filename":"fig4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/9db998efb39a6e9f9ca14746.jpg"},{"id":96650092,"identity":"4cfadf48-2673-4802-9bd8-66b0b73444ae","added_by":"auto","created_at":"2025-11-24 16:07:11","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2506383,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/33664571-f4e0-4540-a55e-1071ce7dd094.pdf"},{"id":86655275,"identity":"8c14f1a1-5fbe-4624-8c80-5fcf158476c1","added_by":"auto","created_at":"2025-07-14 10:11:47","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":226538,"visible":true,"origin":"","legend":"","description":"","filename":"Shodaetal.IVYTrialMSSupplementaryFigures.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/949df3683d141f3d5e42e8b1.pdf"},{"id":86656410,"identity":"4a0247d1-395f-410d-8bb9-7d6035b43b39","added_by":"auto","created_at":"2025-07-14 10:19:47","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":15584,"visible":true,"origin":"","legend":"","description":"","filename":"Shodaetal.IVYTrialMSSupplementaryTables.docx","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/946c04aa875a052f387139fc.docx"},{"id":86655282,"identity":"456b1726-f4fd-41ce-bfb7-db7f6fccfc9e","added_by":"auto","created_at":"2025-07-14 10:11:47","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":34009,"visible":true,"origin":"","legend":"","description":"","filename":"Shodaetal.IVYTrialMSClinicalTrialProtocol.docx","url":"https://assets-eu.researchsquare.com/files/rs-7059542/v1/e021436c5d1c30315b733fa3.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"A machine-learning powered liquid biopsy predicts response to Paclitaxel plus Ramucirumab in advanced gastric cancer: Results from the prospective IVY trial","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eGastric cancer is one of the leading causes of cancer-related deaths worldwide [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Combination therapy with fluoropyrimidine and platinum agents is one of the major first-line treatments for unresectable or metastatic gastric cancer [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. In addition, the prolonged survival benefits from several other regimens, including paclitaxel and ramucirumab (RAM), have also been demonstrated in various randomized trials [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Regarding second-line treatments, the recent RAINBOW trial showed that paclitaxel (PTX) plus RAM exhibited significantly better overall survival (OS) compared to PTX plus placebo in patients with previously treated advanced gastric cancer. Consequently, this treatment regimen is now recognized as the standard second-line chemotherapy [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The latest NCCN guidelines also recommend PTX plus RAM as the standard treatment for second-line chemotherapy, with a response rate of approximately 30% and a disease control rate of approximately 70% [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. However, grade 4 neutropenia and febrile neutropenia have been reported in approximately 25% and 5% of patients who receive this treatment, respectively [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. In other words, ~\u0026thinsp;30% of patients who do not derive any benefit from second-line treatment will not only experience disease progression but will also suffer from the toxicity of such therapy, including a loss of strength and energy. In addition, they may be forced to discontinue or terminate gastric cancer treatment without receiving other treatments that might have provided initial therapeutic efficacy. These patients, who are unable to undergo curative surgery and have transitioned to second-line therapy due to failure of first-line treatment, often face limited survival time. Therefore, selecting the proper treatment in a timely manner is particularly crucial in this setting. The development of clinically useful biomarkers that accurately predict treatment response is essential for guiding therapeutic decisions and improving the clinical management of this population. This highlights the imperative clinical need for the availability of molecular biomarkers that can help predict resistance to second-line therapy and select the subset of gastric cancer patients who will clinically benefit from such a treatment regimen.\u003c/p\u003e\u003cp\u003eIn the context of markers that can predict response to second-line therapy in gastric cancer, the field remains in its infancy. Ideally, considering the biological effects of chemotherapy, such a treatment prediction should be performed in biospecimens obtained prior to the initiation of any treatment. Previous studies have measured tumor tissue levels of vascular endothelial growth factor receptor (VEGFR) and VEGF-related genes, which did not exhibit any predictive potential despite their association with the mechanism(s) of action of RAM [\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Furthermore, predictive biomarkers based on tissue examination are not as attractive due to their invasiveness and the challenges associated with collecting tumor tissues from metastatic tumors before initiating second-line treatment. In this regard, liquid biopsies based on tumor-derived exosomal cargo are emerging rapidly, offering several distinct advantages [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. First, exosomes are extracellular vesicles that facilitate intercellular communication by actively transmitting cellular components, including proteins, DNA, RNA, and various noncoding RNAs [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]; hence, a significant number of tumor-derived molecular signatures can be readily interrogated in exosomes [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Second, the protective shell of the exosomal membranes mitigates the degradation of molecular entities in bodily fluids, thereby dramatically enhancing the stability of exosomal cargo expression [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. These key features, along with the fact that exosome-based liquid biopsies are less invasive and increase patient comfort, are emerging as a paradigm-shifting clinical scenario [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eAccordingly, we undertook a systematic and comprehensive study by performing genomewide transcriptomic profiling of exosomal microRNAs (exo-miRNAs) to help identify and develop predictive biomarkers of response to second-line therapy, analyzing pre-treatment blood specimens from a prospective clinical cohort, the IVY-Trial [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. We successfully established an exo-miRNA-based liquid biopsy assay that can help predict therapeutic efficacy against PTX plus RAM therapy in gastric cancer patients, allowing for a non-invasive and personalized pre-therapeutic selection approach in these patients.\u003c/p\u003e"},{"header":"MATERIALS AND METHODS","content":"\u003cp\u003ePatient cohorts\u003c/p\u003e\u003cp\u003eIn this study, we analyzed a cohort of 162 patients enrolled in a prospective clinical trial (the IVY study) [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. This longitudinal trial was conducted at multiple institutions in Japan as part of the University Hospital Medical Information Network\u0026rsquo;s Clinical Trials Registry (registry number UMIN000033376) and ClinicalTrials.gov registration (NCT06490055) to evaluate the efficacy of second-line treatment in patients with gastric cancer. The eligibility and exclusion criteria were reported previously [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Among this cohort, 115 patients with gastric cancer received PTX plus RAM treatment as second-line chemotherapy between October 1, 2018, and October 31, 2021. Blood samples were collected just prior to second-line therapy, processed for serum isolation, and frozen in a -80\u0026deg;C freezer until use. Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e describes the characteristics of all patients within the clinical cohort. A written informed consent was obtained from all patients, and Institutional Review Board approval was obtained from each participating institution. The study was conducted in compliance with the Declaration of Helsinki.\u003c/p\u003e\u003cp\u003eTreatment methods\u003c/p\u003e\u003cp\u003eThe RAM plus solvent-based (sb)-PTX regimen consisted of RAM (80 mg/m\u0026sup2; administered intravenously on days 1 and 15) and sb-PTX (80 mg/m\u0026sup2; administered intravenously on days 1, 8, and 15) every 4 weeks. RAM plus nanoparticle albumin-bound (nab)-PTX combination therapy comprised of RAM (80 mg/m2 intravenously on days 1 and 15) and nab-PTX (100 mg/m2 intravenously on days 1, 8, and 15) every 4 weeks [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eExosomal RNA extraction from serum specimens\u003c/p\u003e\u003cp\u003eTo prepare libraries for small RNA sequencing, exosomes were first isolated from 400 \u0026micro;L of serum using the exoRNeasy Midi Kit (Qiagen, Valencia, CA), followed by RNA extraction using the miRNeasy Kit (Qiagen, Hilden, Germany). For the real-time quantitative reverse-transcription polymerase chain reaction (qRT-PCR) analysis, total exosomal RNA was isolated from 200 \u0026micro;L of serum using the Total Exosome Isolation Kit (Invitrogen, Waltham, MA) and the miRNeasy kit (Qiagen, Valencia, CA), according to the manufacturer\u0026rsquo;s instructions.\u003c/p\u003e\u003cp\u003eSmall RNA sequencing and data analysis\u003c/p\u003e\u003cp\u003eExosomal small RNA libraries were prepared using the NEXTflex Small RNA-Seq Kit v3 (PerkinElmer, Waltham, MA, USA). The D1000 TapeStation system (RRID: SCR_018435) was used for quality assessment of the sequencing libraries prior to small RNA sequencing on the Illumina NovaSeq platform (RRID: SCR_016387). Subsequently, the raw fastq sequencing data were generated after paired-end sequencing and evaluated using the FastQC tool (RRID: SCR_014583). Next, adapter trimming and low-quality read filtering were performed using the Cutadapt tool (RRID: SCR_011841). Thereafter, miRNAs were aligned and quantified using the miRDeep2 modules with miRBase (version 22, RRID: SCR_01082). Finally, miRNA abundance for each case was normalized based on counts per million (CPM).\u003c/p\u003e\u003cp\u003eReal-time quantitative reverse-transcription polymerase chain reaction assays\u003c/p\u003e\u003cp\u003emiRCURY LNA RT Kit (Qiagen) was used for reverse transcription of RNA to complementary DNA (cDNA). SensiFAST SYBR Lo-ROX Kit (Bioline, London, UK) and QuantStudio 6 Flex RT-PCR System (Applied Biosystems, Foster City, CA, USA, RRID: SCR_020239) were used for performing the qRT-PCR assays. The expression level of target exo-miRNAs was normalized using miR-16-5p by the 2-ΔCt method.\u003c/p\u003e\u003cp\u003eStudy design and evaluation\u003c/p\u003e\u003cp\u003eTumor evaluation by imaging was performed every 2 to 3 months, and treatment response was assessed using the Response Evaluation Criteria in Solid Tumors (RECIST) ver. 1.1 [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Regarding response to second-line therapy, patients who exhibited a partial response (PR) or stable disease (SD) were considered responders. In contrast, those with disease progression (PD) were defined as non-responders. No patient in the study had a complete response to treatment. Progression-free survival (PFS) was defined as the time from enrollment to the first occurrence of progression or death from any cause after initiation of second-line therapy. Overall survival (OS) was defined as the time from enrollment to death or last contact.\u003c/p\u003e\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eDifferential expression analysis of miRNAs between responders and non-responders was performed using the R package 'limma' (version 4.2). Differential miRNA expression (up- and down-regulated miRNAs) was presented as volcano plots using the 'ggplot2' package (version 3.3.6), and heatmaps were generated through the 'pheatmap' package (version 1.0.12). Receiver operating characteristic (ROC) curves to evaluate biomarker performance were established using the 'pROC' package (version 1.18.0). Univariate and multivariate logistic regression were performed using the R package 'DAG' (version 1.25.4).\u003c/p\u003e\u003cp\u003eThe Mann-Whitney U test was used to analyze continuous variables. Survival time analysis included the construction of Kaplan\u0026ndash;Meier survival curves for groups based on univariate predictors, which was performed using the \u0026ldquo;survival\u0026rdquo; and \u0026ldquo;survminer\u0026rdquo; packages in R, with differences assessed via the log-rank test. Multivariate analysis of survival-related clinical factors was performed using the Cox proportional hazards regression model. Nomograms for the Cox hazards regression model were performed using the 'rms' package (version 6.3-0). Statistical significance was determined at a two-tailed P-value threshold of less than 0.05. Statistical analyses were performed using the R statistical software environment (version 4.3.2, RRID: SCR_001905).\u003c/p\u003e\u003c/div\u003e"},{"header":"RESULTS","content":"\u003cp\u003eGenome-wide profiling identified 10 circulating exo-miRNA candidates for differentiating patients with and without response to second-line treatment in gastric cancer\u003c/p\u003e\u003cp\u003eThe primary objective of this study was to identify clinically relevant exo-miRNAs as non-invasive biomarkers for predicting treatment response to PTX plus RAM treatment in gastric cancer. Toward this end, we designed a three-step study that included: i) biomarker discovery by undertaking genome-wide expression profiles of circulating exo-miRNAs that could predict response to PTX plus RAM treatment in gastric cancer, ii) clinical validation for the predictive performance of circulating exo-miRNA biomarker panels, and iii) establishment of a model to predict therapeutic efficacy for routine clinical application (Supplementary Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e and Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). A biomarker discovery effort based on systematic, comprehensive, and unbiased genome-wide small RNA sequencing was initially undertaken using a total of 28 serum exo-miRNA specimens from responders and non-responders treated with second-line therapy.\u003c/p\u003e\u003cp\u003eTo gain an understanding of the circulating exo-miRNA profiles in gastric cancer patients immediately prior to second-line treatment, a comparative analysis was conducted between responders and non-responders. Using the criteria of Log2 Foldchange\u0026thinsp;\u0026gt;\u0026thinsp;1.0 and p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, we identified 19 downregulated and 11 upregulated miRNAs (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). We next narrowed down our miRNA selection by applying the condition of an average expression level [log2 (CPM\u0026thinsp;+\u0026thinsp;1)]\u0026thinsp;\u0026ge;\u0026thinsp;1 and a p-value less than 0.01. Finally, a panel of 10 differentially expressed exo-miRNAs that discriminated between responders and non-responders was identified and prioritized as potential biomarker candidates for further investigations.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSix miRNAs (hsa-miR-10a-5p, hsa-miR-10b-5p, hsa-miR-125a-5p, hsa-miR-139-5p, hsa-miR-335-5p, and miR-379-5p) were significantly upregulated, while the remaining four (hsa-miR-25-5p, hsa-miR-450a-5p, hsa-miR-548ad-5p, and hsa-miR-624-5p), were significantly downregulated in non-responder patients (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB). Ridgeline plots visually depict the extent of distinction between responders and non-responders for each candidate biomarker, utilizing their normalized expression level (Z scores). Importantly, logistic regression analysis has demonstrated that all six upregulated miRNAs and 4 downregulated miRNAs were strongly associated with the response to paclitaxel plus ramucirumab as a second-line treatment in GC (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC and Supplementary Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). The radar chart also visually depicts the absolute coefficients of the logistic regression model for the 10-exo-miRNA panel (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD). In the discovery cohort, the AUC of this 10 exo-miRNA panel was 1.00 (95% CI, 1.00\u0026ndash;1.00; 100% sensitivity, 100% specificity, 100% PPV, 100% NPV, Supplementary Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). Taken together, these biomarker discovery results support our initial hypothesis that circulating exo-miRNA markers may provide superior sensitivity and specificity for predicting response to PTX plus RAM treatment in patients with gastric cancer.\u003c/p\u003e\u003cp\u003ePotential candidate exo-miRNAs involved in gastric cancer progression\u003c/p\u003e\u003cp\u003eTo understand the contribution of our candidate circulating exo-miRNAs to gastric cancer progression, we next investigated their biological significance by identifying the specific growth signaling genes downstream targets for these microRNAs. miRNet 2.0 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.mirnet.ca\u003c/span\u003e\u003cspan address=\"https://www.mirnet.ca\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e], a comprehensive knowledge base that integrates high-quality miRNA-target interaction data, was used to predict binding sites involving candidate miRNAs and information on transcription factors affecting target genes. It was intriguing to note that the target gene clusters for the identified candidate miRNAs revealed a significant involvement not only in the MAPK and Wnt signaling pathways, which are oncogenic pathways involved in gastric cancer progression, but also in the VEGF pathway, the primary mechanistic signaling pathway responsible for RAM (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). Furthermore, the high-risk group, stratified by candidate exo-miRNAs in the discovery cohort, exhibited a significantly poorer prognosis after initiation of second-line therapy compared to the low-risk group (Supplementary Fig. \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e). These results indicate that the candidate exo-miRNAs are not only predictive biomarkers for sensitivity to PTX plus RAM treatment but also potential biomarkers for gastric cancer progression and prognosis after second-line treatment.\u003c/p\u003e\u003cp\u003ePerformance of the exo-miRNA panel for discriminating patients with response to second-line PTX plus RAM treatment in gastric cancer\u003c/p\u003e\u003cp\u003eTo establish a routine clinical predictive assay based on the discovered miRNA biomarkers, qRT-PCR assays were performed on blood specimens obtained from patients in a prospective clinical trial. Information on clinicopathologic factors for this prospective clinical cohort is shown in Supplemental Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e. The expression levels of these 10 candidate exo-miRNAs were examined in a cohort of 115 patients, including 80 responders and 35 non-responders. The results from the PCR assays on the serum specimens were consistent with those in our discovery cohort, indicating that all 10 exo-miRNAs were readily detectable and confirming the robustness of our biomarker discovery. Subsequently, these candidate exo-miRNAs were incorporated into a logistic regression model to train a risk prediction algorithm for identifying the response to PTX plus RAM treatment in patients with gastric cancer. For each of the exo-miRNAs in this panel development, coefficients and constants derived from the logistic regression equation were applied as follows to calculate a risk score for its ability to predict a lack of response to PTX plus RAM therapy: 10-exo-miRNA panel score = (-6.56011 x exo-miR10a) + (-10.48902 x exo-miR10b) + (-205.95086 x exo-miR125a-5p) + (20.62076 x exo-miR139-5p) + (0.32001 x exo-miR25-5p) + (-1.00614 x exo-miR335-3p) + (-0.0024514 x exo-miR379-5p) + (0.76029 x exo-miR450a-5p)\u0026thinsp;+\u0026thinsp;0.023331 x exo-miR548ad) + (-0.097031 x exo-miR624-5p) \u0026minus;\u0026thinsp;0.61411. The combined analysis of these markers with the 10-exo-miRNA panel revealed that this signature was significantly superior in overall predictive accuracy, with the risk score being significantly higher in non-responders compared to responders (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Supplementary Fig. S4A), and a robust corresponding AUC value of 0.83 (Supplementary Fig. S4B). The overall predictive accuracy of this composite panel is consistent with the potential obtained from the discovery cohort study.\u003c/p\u003e\u003cp\u003eThe prognostic potential of the exo-miRNA panel in predicting survival outcomes in patients with second-line treatment in gastric cancer\u003c/p\u003e\u003cp\u003eTo evaluate the prognostic potential of our miRNA panel, we performed a survival analysis by comparing overall survival (OS) and progression-free survival (PFS) in patients categorized as either high- or low-risk based on the 10-exo-miRNA panel. The median follow-up duration for the entire cohort was 18.9 months (95% CI\u0026thinsp;=\u0026thinsp;16.7\u0026ndash;21.4). It was interesting to observe that, in this prospective cohort, high-risk patients exhibited a significantly worse prognosis compared to low-risk patients (PFS: p\u0026thinsp;=\u0026thinsp;0.019; OS: p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Supplementary Fig. S5A, B). These results indicate that, in addition to its ability to predict therapeutic response in gastric cancer patients undergoing second-line therapy, this exo-miRNA panel offers significant prognostic potential in this disease.\u003c/p\u003e\u003cp\u003eEstablishment of a clinically attractive assay for non-invasive identification of therapeutic response prediction to second-line treatment in gastric cancer\u003c/p\u003e\u003cp\u003eSimpler analysis methods are needed for clinical applications, and reducing the number of markers to be analyzed is economically viable. To develop a clinically feasible and cost-effective assay that includes only the minimal number of markers necessary to maintain the overall predictive performance of the exo-miRNAs, a systematic stepwise backward elimination method was used to prioritize the most pertinent biomarker candidates. This statistical strategy resulted in a reduced panel of five exo-miRNA candidates: hsa-miR-10a-5p, hsa-miR-25-5p, hsa-miR-125a-5p, hsa-miR-139-5p, and hsa-miR-450a-5p. The reduced panel of candidate miRNAs was then used to develop a logistic regression equation to recalibrate the final risk prediction model in the clinical cohort as follows: 5-exo-miRNA panel score = (21.31891 x exo-miR10a) + (0.33244 x exo-miR25-5p) + (-315.55693 x exo-miR125a-5p) + (23.60580 x exo-miR139-5p) + (0.60430 x exo-miR450a-5p) \u0026minus;\u0026thinsp;0.60185.\u003c/p\u003e\u003cp\u003eThis recalibrated model was once again applied to the entire clinical cohort which revealed an unchanged predictive performance, with risk scores significantly higher in the higher in the group with a lack of response vs. the response group (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). Notably, the predictive AUC value of this 5-exo-miRNA panel was 0.84, which was higher than the predictive performance of the larger pool of 10 markers (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). Furthermore, the performance of the exo-miRNA panel with this reduced set of markers revealed a higher predictive performance improvement of 88% specificity, indicating the ability to identify gastric cancer patients who will benefit from PTX plus RAM therapy with very high performance (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). Taken together, these results are very promising and highlight that the reduced exo-miRNA panel is a highly robust, clinically attractive, and inexpensive assay for predicting treatment response in gastric cancer patients undergoing PTX plus RAM therapy.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe prognostic potential of the 5-exo-miRNA panel in gastric cancer patients undergoing second-line treatment\u003c/p\u003e\u003cp\u003eTo evaluate the prognostic potential of our recalibrated panel, we performed a survival analysis by comparing OS and PFS in patients categorized as either high or low-risk based on the 5-exo-miRNA panel. High-risk patients had a significantly poorer prognosis than low-risk patients (PFS: p\u0026thinsp;=\u0026thinsp;0.019; OS: p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD, Supplementary Fig. S6). It is worth noting that the median progression-free survival (PFS) for high-risk patients is 1.9 months. In comparison, low-risk patients have a significantly different median PFS of 4.2 months (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). Furthermore, multivariate Cox proportional hazards regression analysis revealed that our 5-exo-miRNA panel was an independent prognostic factor along with body mass index (BMI) in this prospective cohort (HR\u0026thinsp;=\u0026thinsp;2.34, 95% CI\u0026thinsp;=\u0026thinsp;1.56\u0026ndash;4.81, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01; Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). These results reemphasize that this exo-miRNA panel is a powerful prognostic factor as well as has the ability to predict treatment response in gastric cancer patients undergoing second-line therapy.\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\u003eCox proportional hazards regression analysis of overall survival in the IVY study\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e\u003cp\u003eUnivariate\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c5\" namest=\"c4\"\u003e\u003cp\u003eMultivariate\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOR (95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eOR (95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge, \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e\u0026ge;\u003c/span\u003e 65 vs. \u0026lt; 65 years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.98 (0.62\u0026ndash;1.55)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.93\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGender, Male vs. Female\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.77 (0.50\u0026ndash;1.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.24\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBody mass index, \u0026ge; 18 vs. \u0026lt; 18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.50 (0.30\u0026ndash;0.83)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e0.55 (0.33\u0026ndash;0.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e0.02\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrimary tumor location,\u003c/p\u003e\u003cp\u003e Upper third vs. Lower two-thirds\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.65 (0.41\u0026ndash;1.03)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.07\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHistological type, \u003c/p\u003e\u003cp\u003e Diffuse type vs. Intestinal type\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.42 (0.92\u0026ndash;2.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMetastatic site,\u003c/p\u003e\u003cp\u003e Hematogenous vs. Peritoneal or Lymphatic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.09 (0.68\u0026ndash;1.74)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.72\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHistory of primary lesion resection,\u003c/p\u003e\u003cp\u003e Present vs. Absent\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.99 (0.63\u0026ndash;1.54)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.95\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHER2 status, Positive vs. Negative\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e0.63 (0.36\u0026ndash;1.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e0.12\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e5-exo-miRNA panel risk score, High vs. Low\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e2.46 (1.56\u0026ndash;3.86)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e2.34 (1.48\u0026ndash;3.69)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOR, odds ratio, CI, confidence interval.\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eA combination of exosome-based miRNA panel and BMI significantly improved the predictive accuracy of response to second-line therapy in gastric cancer patients\u003c/p\u003e\u003cp\u003eConsidering the current landscape of clinical risk factors involved in the progression of gastric cancer patients with distant metastases, we investigated whether a risk stratification model combining currently used clinicopathological risk factors (i.e., BMI, HER2 status, gender, metastatic site, tumor location of the primary tumor, age, and history of primary tumor resection) with our exo-miRNA panel could further improve predictive accuracy of treatment response in gastric cancer patients receiving second-line therapy. Univariate analysis revealed that BMI (BMI\u0026thinsp;\u0026lt;\u0026thinsp;18; p\u0026thinsp;=\u0026thinsp;0.002) and the 5-exo-miRNA panel (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) were significantly associated with a lack of treatment response in gastric cancer patients receiving PTX plus RAM therapy (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Subsequent multivariate analysis including only significant variables from the univariate model revealed that BMI (BMI\u0026thinsp;\u0026lt;\u0026thinsp;18; p\u0026thinsp;=\u0026thinsp;0.032) and exo-miRNA panel (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001) emerged as independent risk factors for a lack of response to PTX plus RAM treatment. (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eWe next examined whether the robustness of our 5-exo-miRNA panel could be further improved by including BMI, which has been independently found to be associated with a lack of treatment response in multivariate analysis. In support of our hypothesis, we found that combining the exo-miRNA panel with BMI improved the AUC from 0.84 to 0.87, thus significantly improving the robustness of this clinical cohort in predicting a lack of response to treatment. (Risk model AUC, 0.87; 95% CI, 0.80\u0026ndash;0.93) (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). We defined this composite signature score as EXEMPLAR (EXo-miRNAs-based Model for Predicting Response to PacLitAxel plus Ramucirumab Treatment). EXEMPLAR was significantly superior in overall predictive accuracy, with risk scores significantly higher in non-responders vis-\u0026agrave;-vis responders (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). The detailed distribution of EXEMPLAR risk scores in gastric cancer patients is shown in the waterfall plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD). The waterfall plot analysis also depicts that EXEMPLAR successfully discriminated between responders and non-responders with a sensitivity of 86%, specificity of 82%, and a true negative of 65/70 (92.9%). These findings again highlight that, although this exosome-based miRNA assay was quite robust on its own, combining it with BMI results in a significant improvement in overall predictive accuracy, emphasizing its potential for translation into the clinic for predicting patients who lack a response to PTX plus RAM treatment.\u003c/p\u003e\u003cp\u003eAn exosome-based liquid biopsy assay offers significant advantages for predicting therapeutic outcomes in patients with gastric cancer undergoing second-line treatment.\u003c/p\u003e\u003cp\u003eIn current clinical practice, there are no standardized tests or biomarkers that can predict the efficacy of PTX plus RAM therapy for gastric cancer patients prior to treatment. Therefore, the standard second-line chemotherapy, RAM plus PTX, is recommended for all patients who have failed to respond to first-line therapy or who are unable to continue treatment due to various side effects. Although PTX plus RAM is undoubtedly a proper regimen that has shown some efficacy in clinical trials, the cytotoxic nature of anticancer therapy may not be helpful for the 30% of patients who receive treatment but will not benefit from it. Therefore, the clinical utility of a treatment strategy should be evaluated by weighing the benefits against the harms and therapeutic effects. To further investigate the clinical significance of our liquid biopsy assay, we performed a decision curve analysis. As shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD, the x-axis represents the threshold probability for a lack of response to PTX plus RAM treatment, and the y-axis represents the net benefit achieved. The decision curve analysis revealed that EXEMPLAR achieved a higher net benefit score across most threshold probabilities compared to PTX plus RAM treatment for all patients, and no patients received PTX plus RAM treatment (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE).\u003c/p\u003e\u003cp\u003eFor example, at a threshold probability of 0.30, liquid biopsy with EXEMPLAR achieved a significantly higher net benefit of about 0.18 for PTX plus RAM treatment. In contrast, the net benefit based on the intervention strategy for all patients was significantly lower, at almost zero. These findings suggest that, in terms of disadvantage avoidance, EXEMPLAR provides significantly higher clinical benefits compared to interventions in all patients or no intervention in any patients.\u003c/p\u003e\u003cp\u003eEstablishment of an EXEMPLAR-based risk nomogram in gastric cancer patients undergoing second-line treatment for a personalized therapeutic approach\u003c/p\u003e\u003cp\u003eSubsequently, we evaluated the impact of the EXEMPLAR, which was established using the statistically significant factors exo-miRNA panel and BMI, on the prognostic value in patients undergoing second-line treatment for gastric cancer. EXEMPLAR-high patients had a significantly poorer prognosis in both OS and PFS compared to low-risk patients (OS: p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; PFS: p\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA, B). The details of PFS and OS for each case are shown in a swimmer plot in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC, and the survival analysis for EXEMPLAR-low and EXEMPLAR-high patients is summarized in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. It should be noted that the median PFS for EXEMPLAR-low and EXEMPLAR-high patients was 4.8 and 1.9 months, respectively, a difference of more than 2-fold. Regarding OS, it is also noteworthy that EXEMPLAR-high patients had no 3-year survivors after 2nd-line initiation of therapy, whereas EXEMPLAR-low patients achieved a 23% 3-year survival rate.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSurvival according to EXEMPLAR status\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eOverall\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eEXEMPLAR high\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eEXEMPLAR low\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003eProgression-free survival\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e3-month PFS rate, %\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e60.0%\u003c/p\u003e\u003cp\u003e(50.9\u0026ndash;68.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e28.9%\u003c/p\u003e\u003cp\u003e(17.7\u0026ndash;43.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e80.0%\u003c/p\u003e\u003cp\u003e(69.2\u0026ndash;87.7%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e6-month PFS rate, %\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25.2%\u003c/p\u003e\u003cp\u003e(18.2\u0026ndash;33.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e11.1%\u003c/p\u003e\u003cp\u003e(4.8\u0026ndash;23.4%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e34.3%\u003c/p\u003e\u003cp\u003e(24.3\u0026ndash; 46.0%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian PFS, months\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e3.97\u003c/p\u003e\u003cp\u003e(3.00\u0026ndash;4.67)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.93\u003c/p\u003e\u003cp\u003e(1.27\u0026ndash;2.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4.81\u003c/p\u003e\u003cp\u003e(4.10\u0026ndash;5.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHazard Ratio of Progression\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.12\u003c/p\u003e\u003cp\u003e(1.44\u0026ndash;3.12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"4\" nameend=\"c4\" namest=\"c1\"\u003e\u003cp\u003eOverall Survival\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e1-year OS rate, %\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e67.8%\u003c/p\u003e\u003cp\u003e(58.8\u0026ndash;75.7%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e55.6%\u003c/p\u003e\u003cp\u003e(41.2\u0026ndash;69.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e75.7%\u003c/p\u003e\u003cp\u003e(64.5\u0026ndash;84.2%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e2-year OS rate, %\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e22.6%\u003c/p\u003e\u003cp\u003e(15.9\u0026ndash;31.1%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e8.9%\u003c/p\u003e\u003cp\u003e(3.5\u0026ndash;20.7%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e31.4%\u003c/p\u003e\u003cp\u003e(21.8\u0026ndash;43.0%)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMedian OS, months\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e1.4\u003c/p\u003e\u003cp\u003e(1.2\u0026ndash;1.5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e1.1\u003c/p\u003e\u003cp\u003e(0.9\u0026ndash;1.3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e1.7\u003c/p\u003e\u003cp\u003e(1.4\u0026ndash;1.9)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;0.01\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHazard Ratio of Death\u003c/p\u003e\u003cp\u003e(95% CI)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e2.56\u003c/p\u003e\u003cp\u003e(1.67\u0026ndash;3.92)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eReference\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"5\" nameend=\"c5\" namest=\"c1\"\u003e\u003cp\u003e* P-values refer to comparisons between the Exo-miRs model high vs. low, CI, confidence interval.\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eFinally, to facilitate the clinical translation of our findings, we developed a prognostic risk nomogram for PFS. These results indicated that our exo-miRNA panel score carried a higher weight than BMI, once again highlighting its clinical significance for an expedient translation into clinical practice (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). Taken together, these data suggest that combining our exo-miRNA panel with the nutritional indicator BMI could provide a more sophisticated personalized medicine approach by proposing a prognostic risk probability model for GC patients undergoing PTX plus RAM treatment.\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eWith today's advances in cancer treatment and the increasing availability of various treatment options, it is essential to develop treatment strategies that provide optimal care for each patient [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Many drugs have been developed for the treatment of patients with advanced gastric cancer, and their efficacy has been demonstrated in clinical trials [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. We have entered an era in which long-term survival is possible even for patients with advanced gastric cancer if they respond to chemotherapy [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. However, the prognosis is still poor for patients with advanced gastric cancer who are untreated or do not respond to treatment [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Therefore, the importance of selecting second-line treatment is evident, especially for patients who have failed to respond to first-line therapy, and establishing a predictive risk model for treatment efficacy is undoubtedly desirable. Our study is a step in this direction, and we have successfully identified a novel exo-miRNAs model that robustly identifies treatment response in gastric cancer patients undergoing second-line therapy through systemic and comprehensive biomarker discovery and validation. More interestingly, our established exo-microRNA model was also able to individually stratify the prognosis of gastric cancer patients undergoing treatment after initiation of therapy. These results highlight the potential clinical significance of a novel exo-miRNAs model for predicting treatment response in gastric cancer patients undergoing second-line therapy. To date, there are no established predictive biomarkers for PTA plus RAM treatment, the main second-line treatment for gastric cancer.\u003c/p\u003e\u003cp\u003e For patients with gastric cancer, the NCCN guidelines recommend the use of fluoropyrimidine and platinum-based chemotherapy as one of the main first-line therapies, with second-line therapy for patients who do not respond to first-line therapy or who have difficulty continuing treatment. PTX plus RAM therapy is the primary second-line treatment, based on the results of the Phase III RAINBOW and REGARD trials [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Other treatment options include several immune checkpoint inhibitors, irinotecan, trifluridine, and other molecularly targeted agents such as trastuzumab deruxtecan and zolbetuximab [\u003cspan additionalcitationids=\"CR22 CR23 CR24\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. However, persistent lack of response to treatment may, in some cases, necessitate termination of treatment during the first or second line of treatment. For patients who fail to respond to first-line treatment, failure to respond to second-line treatment also results in a loss of strength and energy, forcing patients to discontinue or terminate treatment without receiving other treatments that might have been effective. These realities in clinical practice highlight the need to develop tests and markers for predicting the effectiveness of second-line treatment. Therefore, several previous studies have attempted to find biomarkers to predict the therapeutic efficacy of PTX plus RAM therapy [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. RAINBOW investigators attempted to develop a biomarker targeting VEGF, the primary mechanism of action of RAM, and examined the association between plasma levels of VEGF-D and therapeutic response, but were unable to demonstrate a significant therapeutic predictive marker effect of ramucirumab in gastric cancer [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Lorenzo et al. suggested that an early increase in soluble VEGFR-2 after one cycle of treatment may be a potential predictive biomarker, and Raghav et al. developed a gene signature using a random forest machine learning model with resected tissue for predicting PTX efficacy [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. Ideally, however, the patient's condition should be assessed in real-time immediately prior to the start of treatment, given the anticipated impact of primary treatment on the outcome of second-line treatment. In this regard, it is worth noting that our study is prospective, in which blood was collected immediately before treatment for a specific regimen. To our knowledge, this is the first report on a biomarker that predicts treatment response for PTX plus RAM using liquid biopsies from a prospective study conducted just prior to treatment.\u003c/p\u003e\u003cp\u003eWe employed a systematic and comprehensive biomarker discovery approach to develop a blood-based exo-miRNA panel with superior performance in predicting treatment outcomes in second-line gastric cancer therapy. Interestingly, in a public database analysis, these exo-miRNA biomarkers were shown to be associated with the VEGF pathway, the primary mechanism of action of RAM, supporting our finding that RAM-based therapeutic regimens correlate with therapeutic efficacy [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Furthermore, our exo-miRNA panel demonstrated high predictive ability in a prospective clinical cohort, clearly stratifying patient prognosis after treatment. In addition, other clinical factors correlated with treatment response were analyzed using multivariate logistic regression analysis, and BMI was selected in conjunction with our exo-miRNA panel.\u003c/p\u003e\u003cp\u003eIn gastric cancer, high pre-treatment BMI and muscle mass are significantly associated with better prognosis after treatment, and nutritional status during chemotherapy has also been reported to correlate with oncologic prognosis28. Conversely, patients with sarcopenia have an increased incidence of treatment-induced side effects, which are also thought to be associated with poor treatment adherence and prognosis [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. These phenomena are considered one of the characteristics of gastric cancer treatment and are called BMI paradoxes [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. An interesting aspect of this study is that a more robust predictive model was established by combining the exo-miRNA panel, which was explored in relation to tumor molecular biology, with BMI, a patient's nutritional index.\u003c/p\u003e\u003cp\u003eFrom a clinical perspective, our exo-miRNAs model has the potential to enable several important improvements when introduced into clinical practice. (i) The possibility of selecting a treatment alternative to PTX plus RAM for patients expected to be resistant to treatment can be considered. (ii) For patients who are predicted to respond well to treatment, patients and physicians can make treatment choices based on a solid perception of the benefits of treatment, even if some risk of chemotherapy side effects is expected. (iii) For patients receiving first-line therapy who are expected to respond well to second-line therapy, early conversion to second-line therapy may be recommended, given the irreversible side effects of primary therapy. Furthermore, our liquid biopsy-based exo-miRNA model is non-invasive and repeatable, making it potentially adaptable for assessing risk status at multiple time points throughout a patient's clinical course. Its performance in this regard should be examined in future evaluations of multiple liquid biopsies in the same patient.\u003c/p\u003e\u003cp\u003eWe would like to acknowledge a few potential limitations of our present study. First, although our study was a prospective clinical study focusing on blood immediately before second-line treatment of gastric cancer, prospective studies using larger patient cohorts are also needed before these biomarkers can be examined in actual clinical practice. Second, real-time evaluation of exosomes may enable the monitoring of cancer progression, and evaluating pre- and post-treatment exosome expression in the same patient may be promising. Nevertheless, our current study provides promising evidence for the clinical significance of the exo-miRNA model in predicting treatment response in gastric cancer patients undergoing second-line treatment, marking an important step toward the application of robust molecular biomarkers for risk assessment and management of lethal malignancies, such as advanced metastatic gastric cancer.\u003c/p\u003e\u003cp\u003eIn conclusion, we have successfully validated noninvasive liquid biopsies in a prospective clinical cohort using genome-wide exo-miRNAs expression profiling to predict treatment response in gastric cancer patients receiving second-line therapy. This is clinically crucial for patients with advanced gastric cancer to develop an optimized treatment plan. Furthermore, our assay provides a framework for a personalized medicine approach, enabling risk-based management and treatment tailored to each patient prior to the initiation of therapy.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eEthics approval and consent to participate :\u0026nbsp;This study was conducted in accordance with the Declaration of Helsinki. All procedures were approved by the Ethics Committee of Kawasaki Medical School. Written informed consent was obtained from all participants prior to enrollment in the study.\u003c/p\u003e\n\u003cp\u003eConsent for publication:\u0026nbsp;Not applicable.\u003c/p\u003e\n\u003cp\u003eAvailability of data and materials:\u0026nbsp;The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003eCompeting interests:\u0026nbsp;The authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003eFunding:\u0026nbsp;This work was supported by grants CA72851, CA181572, CA187956, CA227602, CA214254, and CA271443 from the National Cancer Institute, NIH.\u003c/p\u003e\n\u003cp\u003eAuthors\u0026apos; contributions:\u0026nbsp;Study concept and design: KS, YA, AG. Provision of samples and clinical data: KS, TN, AG. Bioinformatic analysis: CX. Data collection and analysis: KS, CX, YA, AM. Drafting of the manuscript: KS, CX, YA, TN, DI, AG. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003eAcknowledgements: The authors thank Drs. Yoh Asahi and Alessandro Mannucci for their thoughtful discussions and advice during this project.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eSung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-249.\u003c/li\u003e\n\u003cli\u003eAjani JA, D\u0026apos;Amico TA, Bentrem DJ, et al. Gastric Cancer, Version 2.2022, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw 2022;20:167-192.\u003c/li\u003e\n\u003cli\u003eFuchs CS, Tomasek J, Yong CJ, et al. Ramucirumab monotherapy for previously treated advanced gastric or gastro-oesophageal junction adenocarcinoma (REGARD): an international, randomised, multicentre, placebo-controlled, phase 3 trial. Lancet 2014;383:31-39.\u003c/li\u003e\n\u003cli\u003eWilke H, Muro K, Van Cutsem E, et al. Ramucirumab plus paclitaxel versus placebo plus paclitaxel in patients with previously treated advanced gastric or gastro-oesophageal junction adenocarcinoma (RAINBOW): a double-masked, randomised phase 3 trial. Lancet Oncol 2014;15:1224-35.\u003c/li\u003e\n\u003cli\u003eXu RH, Zhang Y, Pan H, et al. Efficacy and safety of weekly paclitaxel with or without ramucirumab as second-line therapy for the treatment of advanced gastric or gastroesophageal junction adenocarcinoma (RAINBOW-Asia): a randomised, multicentre, double-masked, phase 3 trial. Lancet Gastroenterol Hepatol 2021;6:1015-1024.\u003c/li\u003e\n\u003cli\u003eVan Cutsem E, Muro K, Cunningham D, et al. Biomarker analyses of second-line ramucirumab in patients with advanced gastric cancer from RAINBOW, a global, randomized, double-blind, phase 3 study. Eur J Cancer 2020;127:150-157.\u003c/li\u003e\n\u003cli\u003eNatsume M, Shimura T, Iwasaki H, et al. Placental growth factor is a predictive biomarker for ramucirumab treatment in advanced gastric cancer. Cancer Chemother Pharmacol 2019;83:1037-1046.\u003c/li\u003e\n\u003cli\u003eAprile G, Ongaro E, Del Re M, et al. Angiogenic inhibitors in gastric cancers and gastroesophageal junction carcinomas: A critical insight. Crit Rev Oncol Hematol 2015;95:165-78.\u003c/li\u003e\n\u003cli\u003eShigeyasu K, Toden S, Zumwalt TJ, et al. Emerging Role of MicroRNAs as Liquid Biopsy Biomarkers in Gastrointestinal Cancers. Clin Cancer Res 2017;23:2391-2399.\u003c/li\u003e\n\u003cli\u003eYu W, Hurley J, Roberts D, et al. Exosome-based liquid biopsies in cancer: opportunities and challenges. Ann Oncol 2021;32:466-477.\u003c/li\u003e\n\u003cli\u003eTang XH, Guo T, Gao XY, et al. Exosome-derived noncoding RNAs in gastric cancer: functions and clinical applications. Mol Cancer 2021;20:99.\u003c/li\u003e\n\u003cli\u003eHuang T, Song C, Zheng L, et al. The roles of extracellular vesicles in gastric cancer development, microenvironment, anti-cancer drug resistance, and therapy. Mol Cancer 2019;18:62.\u003c/li\u003e\n\u003cli\u003eKalluri R, LeBleu VS. The biology, function, and biomedical applications of exosomes. Science 2020;367.\u003c/li\u003e\n\u003cli\u003eTanioka H, Nagasaka T, Uno F, et al. The relationship between peripheral neuropathy and efficacy in second-line chemotherapy for unresectable advanced gastric cancer: a prospective observational multicenter study protocol (IVY). BMC Cancer 2019;19:941.\u003c/li\u003e\n\u003cli\u003eShitara K, Takashima A, Fujitani K, et al. Nab-paclitaxel versus solvent-based paclitaxel in patients with previously treated advanced gastric cancer (ABSOLUTE): an open-label, randomised, non-inferiority, phase 3 trial. Lancet Gastroenterol Hepatol 2017;2:277-287.\u003c/li\u003e\n\u003cli\u003eBando H, Shimodaira H, Fujitani K, et al. A phase II study of nab-paclitaxel in combination with ramucirumab in patients with previously treated advanced gastric cancer. Eur J Cancer 2018;91:86-91.\u003c/li\u003e\n\u003cli\u003eEisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 2009;45:228-47.\u003c/li\u003e\n\u003cli\u003eFan Y, Siklenka K, Arora SK, et al. miRNet - dissecting miRNA-target interactions and functional associations through network-based visual analysis. Nucleic Acids Res 2016;44:W135-41.\u003c/li\u003e\n\u003cli\u003eJoshi SS, Badgwell BD. Current treatment and recent progress in gastric cancer. CA Cancer J Clin 2021;71:264-279.\u003c/li\u003e\n\u003cli\u003eAlsina M, Arrazubi V, Diez M, et al. Current developments in gastric cancer: from molecular profiling to treatment strategy. Nat Rev Gastroenterol Hepatol 2023;20:155-170.\u003c/li\u003e\n\u003cli\u003eKang YK, Chen LT, Ryu MH, et al. Nivolumab plus chemotherapy versus placebo plus chemotherapy in patients with HER2-negative, untreated, unresectable advanced or recurrent gastric or gastro-oesophageal junction cancer (ATTRACTION-4): a randomised, multicentre, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2022;23:234-247.\u003c/li\u003e\n\u003cli\u003eShitara K, Van Cutsem E, Bang YJ, et al. Efficacy and Safety of Pembrolizumab or Pembrolizumab Plus Chemotherapy vs Chemotherapy Alone for Patients With First-line, Advanced Gastric Cancer: The KEYNOTE-062 Phase 3 Randomized Clinical Trial. JAMA Oncol 2020;6:1571-1580.\u003c/li\u003e\n\u003cli\u003eHironaka S, Ueda S, Yasui H, et al. Randomized, open-label, phase III study comparing irinotecan with paclitaxel in patients with advanced gastric cancer without severe peritoneal metastasis after failure of prior combination chemotherapy using fluoropyrimidine plus platinum: WJOG 4007 trial. J Clin Oncol 2013;31:4438-44.\u003c/li\u003e\n\u003cli\u003eShitara K, Doi T, Dvorkin M, et al. Trifluridine/tipiracil versus placebo in patients with heavily pretreated metastatic gastric cancer (TAGS): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2018;19:1437-1448.\u003c/li\u003e\n\u003cli\u003eShitara K, Bang YJ, Iwasa S, et al. Trastuzumab Deruxtecan in Previously Treated HER2-Positive Gastric Cancer. N Engl J Med 2020;382:2419-2430.\u003c/li\u003e\n\u003cli\u003eFornaro L, Musettini G, Orlandi P, et al. Early increase of plasma soluble VEGFR-2 is associated with clinical benefit from second-line treatment of paclitaxel and ramucirumab in advanced gastric cancer. Am J Cancer Res 2022;12:3347-3356.\u003c/li\u003e\n\u003cli\u003eSundar R, Barr Kumarakulasinghe N, Huak Chan Y, et al. Machine-learning model-derived gene signature predictive of paclitaxel survival benefit in gastric cancer: results from the randomised phase III SAMIT trial. Gut 2022;71:676-685.\u003c/li\u003e\n\u003cli\u003eMeng Q, Tan S, Jiang Y, et al. Post-discharge oral nutritional supplements with dietary advice in patients at nutritional risk after surgery for gastric cancer: A randomized clinical trial. Clin Nutr 2021;40:40-46.\u003c/li\u003e\n\u003cli\u003eKawamura T, Makuuchi R, Tokunaga M, et al. Long-Term Outcomes of Gastric Cancer Patients with Preoperative Sarcopenia. Ann Surg Oncol 2018;25:1625-1632.\u003c/li\u003e\n\u003cli\u003eKamiya H, Komatsu S, Nishibeppu K, et al. Obesity paradox as a new insight from postoperative complications in gastric cancer. Sci Rep 2023;13:10116.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"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":"molecular-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"molc","sideBox":"Learn more about [Molecular Cancer](http://gsejournal.biomedcentral.com/)","snPcode":"12943","submissionUrl":"https://submission.nature.com/new-submission/12943/3","title":"Molecular Cancer","twitterHandle":"@SN_Oncology","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Gastric cancer, Liquid biopsy, Exosomal microRNAs, Ramucirumab, Paclitaxel, Machine learning, Biomarker","lastPublishedDoi":"10.21203/rs.3.rs-7059542/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7059542/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePaclitaxel plus ramucirumab (PTX + RAM) is a widely used second-line treatment for advanced gastric cancer, yet no validated biomarkers exist to predict therapeutic response. Identifying non-invasive predictors could enable patient stratification and optimize outcomes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted a prospective observational multicenter study (IVY trial; NCT06490055) enrolling 115 patients with advanced gastric cancer treated with PTX + RAM. Serum was collected prior to the initiation of treatment. Small RNA sequencing identified differentially expressed exosomal microRNAs (exo-miRNAs) between responders and non-responders. Machine learning and logistic regression were employed to construct a predictive model, which was subsequently validated using quantitative real-time polymerase chain reaction (qRT-PCR) in the entire cohort.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTen candidate exo-miRNAs were initially discovered, and a five-miRNA panel (miR-10a-5p, miR-25-5p, miR-125a-5p, miR-139-5p, and miR-450a-5p) was selected via stepwise elimination. This 5-exo-miRNA model achieved high accuracy in distinguishing responders from non-responders (AUC = 0.84). When combined with body mass index (BMI), the composite model (EXEMPLAR) demonstrated enhanced predictive performance (AUC = 0.87). High-risk patients exhibited significantly shorter progression-free survival (PFS: median, 1.9 vs. 4.2 months, p = 0.019) and overall survival (OS: median, 1.1 vs. 1.7 years, p \u0026lt; 0.001). Decision curve analysis confirmed the clinical benefit of the model. A nomogram was developed to facilitate personalized risk assessment.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study identifies and validates a novel 5-exo-miRNA panel for predicting response to second-line PTX plus RAM therapy in gastric cancer. The combined exosomal signature and BMI risk model provides a clinically applicable, non-invasive tool for personalized treatment selection.\u003c/p\u003e\n\u003cp\u003eClinicalTrials.gov Identifier: NCT06490055\u003c/p\u003e","manuscriptTitle":"A machine-learning powered liquid biopsy predicts response to Paclitaxel plus Ramucirumab in advanced gastric cancer: Results from the prospective IVY trial","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-14 10:11:39","doi":"10.21203/rs.3.rs-7059542/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-08-20T16:46:19+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-17T10:34:05+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-08-06T14:49:33+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"68467144484122303926990185350369676912","date":"2025-07-30T17:50:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"305617442335404558251764586281390483573","date":"2025-07-29T13:51:41+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-29T08:56:57+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"181957406735017971680039197774153283959","date":"2025-07-29T05:38:27+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"292536342513067792110300952922786265097","date":"2025-07-29T03:14:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"114107326584785255928111431532854549451","date":"2025-07-18T02:00:22+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-07-08T12:01:08+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-07T04:38:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-07T04:37:21+00:00","index":"","fulltext":""},{"type":"submitted","content":"Molecular Cancer","date":"2025-07-06T18:52:03+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"molecular-cancer","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"molc","sideBox":"Learn more about [Molecular Cancer](http://gsejournal.biomedcentral.com/)","snPcode":"12943","submissionUrl":"https://submission.nature.com/new-submission/12943/3","title":"Molecular Cancer","twitterHandle":"@SN_Oncology","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"195ea208-fc04-47ef-b386-0852399fb886","owner":[],"postedDate":"July 14th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-11-24T16:00:54+00:00","versionOfRecord":{"articleIdentity":"rs-7059542","link":"https://doi.org/10.1186/s12943-025-02526-0","journal":{"identity":"molecular-cancer","isVorOnly":false,"title":"Molecular Cancer"},"publishedOn":"2025-11-22 15:56:59","publishedOnDateReadable":"November 22nd, 2025"},"versionCreatedAt":"2025-07-14 10:11:39","video":"","vorDoi":"10.1186/s12943-025-02526-0","vorDoiUrl":"https://doi.org/10.1186/s12943-025-02526-0","workflowStages":[]},"version":"v1","identity":"rs-7059542","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7059542","identity":"rs-7059542","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}